Open main menu

User talk:Andrewa/P T test cases

Why this pageEdit

This is for particularly interesting cases raised at User talk:Andrewa/P T examples and scenarios. Andrewa (talk) 01:02, 5 July 2019 (UTC)

Of course the purpose of both pages is to support User:Andrewa/Primary Topic RfC. Andrewa (talk) 00:36, 9 July 2019 (UTC)


Various nationalist groups lay claim to the name Macedonia. The DAB is currently at the base name, which avoids the POV of siding with any of them.

But siding with none of them is also a POV. And under current policy we are saying, in Wikipedia's voice, that they are all wrong.

Far better to avoid doing this by just saying instead, the term is ambiguous, it means different things to different people, so for that reason alone we should not use it as the article title for any of the possible topics.

It may be in some sense proper to use it for one of the several topics (that's what we mean by Primary Topic), but to do so is POV.

But more subtly, it's also POV to consider using it for one of these topics, and then decide there's not enough evidence to do so. The various partisans all think they have such claims. We should not deny any of these claims in Wikipedia's voice. And currently we deny all of them.

If instead we just say that these article titles should be unambiguous, there is no such problem. We need not then take any side at all.

See Talk:Macedonia for some heated POV discussion, and Wikipedia:Requests for arbitration/Eastern Europe for some equally involved ARBCOM activity. See User:Andrewa/P T test cases#Macedonia for links to at least some of the many other previous discussions, and please add any I have missed. Andrewa (talk) 22:12, 26 July 2019 (UTC)

Applying WP:PRIMARYTOPIC criteria the same way we apply it to any other ambiguous name is deciding this with a NPOV. That assumes we have a way to determine the most likely sought target when someone searches with "Macedonia". --В²C 23:24, 26 July 2019 (UTC)
Thanks for that, it illustrates the point very well.
No, it's not NPOV, because it assumes more than that. It assumes that the only criteria we're applying are objective ones, such as page views. But we've failed to come up with such a policy, despite many attempts (mostly by yourself). And one of the reasons for that failure is the strong consensus that page views do not always tell the whole story with respect to Primary Topic. See User:Andrewa/The Problem With Page Views#Examples.
So the significance criterion stays in for the moment, and there is no prospect of removing it. (And were we to do so the results would IMO be an unworkable mess, even if we were to agree on criteria for avoiding recentism, and without these it would be even more of a mess.)
And with that guideline in place, we are saying, in Wikipedia's voice, that none of the various claims for Primary Topic meet that guideline... that is, that none of the uses are of sufficient significance to warrant making that use the Primary Topic. The various claimants all have POVs that their usage is of such significance, and we are denying all such claims.
And that may even be a good thing to do! But doing it in Wikipedia's voice is not a good thing, and violates one of our most fundamental policies. Andrewa (talk) 17:21, 28 July 2019 (UTC)

The AmericansEdit

The Americans (2013 TV series) was moved to The Americans after much discussion, during which it was claimed that arriving instead at The Americans (disambiguation) was either as bad as, or perhaps even worse than, arriving at the wrong article. Andrewa (talk) 06:24, 5 July 2019 (UTC)

Amakuru has commented elsewhere In the case of "The Americans", a dab would indeed seem best. But that doesn't mean primary topics are wrong in a more general sense. It's an unfair poster boy for the point, because it's genuinely ambiguous. Cheers. [1]

If we had consensus support for the DAB being best, IMO that would be reason enough that it is a very good poster boy. It would mean that we had decided by consensus that, after extensive discussion, a highly respected admin has performed a move that has damaged Wikipedia. [2] Isn't that a rather spectacular failure of the current policy?

That is an interesting point, but I think the various predictions that while it's the Primary Topic now, it's unlikely to remain so, are even more interesting. There is even speculation that it may lose its Primary Topic Status, but then regain it briefly in a couple of decades, and then lose it again. We must be careful of wp:BALL of course, but is such a fickle criterion really what we want as the basis for disambiguating article names? Andrewa (talk) 11:03, 9 July 2019 (UTC)

Genuinely ambiguousEdit

It's an unfair poster boy for the point, because it's genuinely ambiguous. (diff above)

Possibly, genuinely ambiguous here means, there's no Primary Topic. That's another way of approaching the problem which will lead to the same conclusion IMO. In the sense that some will always disagree as to what the Primary Topic is (and this is shown by the failure to find any objective measure for P T despite many attempts) there is never a Primary Topic for an ambiguous term (but note the qualification In the sense...). That is, all ambiguous article names are genuinely ambiguous, in any objective sense. Andrewa (talk) 19:10, 10 July 2019 (UTC)

What is the harmEdit

More discussion on this at User talk:Andrewa/P T examples and scenarios#The Americans. Andrewa (talk) 23:20, 12 July 2019 (UTC)

This seems to be the place to discuss the harm of having the article on The Americans (2013 TV Series) at the base name The Americans as presently following this RM close. In no particular order.

  • Attempting to Wikilink to any of the seven articles listed at The Americans (disambiguation) by simply typing [[The Americans]] gives no warning that the term is ambiguous, and if any of the other six meanings is intended, a mislinking will result.
  • Bookmarking or otherwise creating an incoming external link to will create a link that is likely to be broken by a future article move.
  • Search engines may not include the redir from The Americans (2013 TV series) in their results lists, so a user searching for this article on one of these but aware of other meanings of The Americans may not find the article at all.

Probably more to follow. Andrewa (talk) 00:07, 14 July 2019 (UTC)

Born2cycle wrote Proposal would make things worse for the vast majority of users using the search term "The Americans" in WP GO Search (others are unaffected and so irrelevant to the question of whether this change makes things better or worse for them). [3]

Others are unaffected and so irrelevant? How about the issues raised above? Others are affected, the only question is how badly.

And the only reason it doesn't make things worse for all users is the resulting redirect left by the move. As commented elsewhere, if we must have articles at ambiguous names, we should at least have a redirect from an unambiguous name for all of these. Andrewa (talk) 21:06, 26 July 2019 (UTC)

Further on this, have a look at #How the dropdown works NOT. Assuming that right now most users typing The Americans into the search box want the article on the 2013 series (now at the base name), it would be better for them to have the series at an unambiguous name. Andrewa (talk) 23:56, 5 August 2019 (UTC)

The Americans disputedEdit

Predictably, this example has been disputed [4].

The result doesn't make much sense to me, but makes some points which need discussion.

Born2cycle, I've done a refactor to make it clear which are my opinions and which are yours. Feel free to further update what is now your own section, so that it best represents your views. Andrewa (talk) 20:41, 14 July 2019 (UTC)

  • Commercial products in popular culture make for really bad cases studies for informing policy discussions, as they are rife with subtle fandom-related POV issues and similar biases. --SmokeyJoe (talk) 04:01, 15 July 2019 (UTC)
    • SmokeyJoe, even if all uses of a given term are "commercial products in popular culture" and one of the uses is far more likely to be sought than the others? If so, then this view flies in the face of community consensus about every example I can think of off the top of my head... Superman, The Sopranos, The Godfather, Law & Order, Emergency!, Occupied, ... --В²C 19:35, 15 July 2019 (UTC)
    • But the policy needs to deal with these and similar cases. So such cases need to be discussed. Andrewa (talk) 22:35, 15 July 2019 (UTC)
      • I think B2C obsesses with pageviews to the exclusion of sensible structure for an encyclopedia. However, this is not to say that pageviews don't matter. The Americans, Superman, The Sopranos, The Godfather, Law & Order are extraordinary in terms of page views, international recognition, as well as *obviously non-trivial scholarly significance*, which with SMALLDETAILS justified their current titles, presumably forever as well. I don't dispute The Americans for this reason, among other reasons. Emergency! I have ore trouble with, with the SMALLDETAIL being terminating character that is a sentence terminating character. The period (".") and the comma (",") are obviously too-small a SMALLDETAIL as final characters in a title, but the terminating "!" is at the boundary. Also, Emergency!, despite massive significance at the time, is now becoming old and unknown to most readers, and while it has massive US 1970s significance and was internationally syndicated, it was syndicated under other titles. In short, massively popular US TV shows should be considered special cases, and the principle doesn't extrapolate to this month's new album even if it does have an abundance of pageviews this month.
I think maybe a better measure than page views ratios between competing topics is absolute page views. I have long argued that an important underappreciated underdocumented criteria against PT is propensity for mis-recognition by large particular audiences (eg school children in India).
When it comes to commercial products, the significance and pageviews should be expected to have a higher threshold, as the owners/promoters are deliberately trying to confuse, and will succeed. On the other hand, a moderately well known scientific or historical cultural topic should not be expected to have high pageviews to displace a topic as PT for a PRECISE-failing short basename. --SmokeyJoe (talk) 00:52, 16 July 2019 (UTC)
Excellent points. These are good reasons for the significance criterion, and will be a problem with any other "objective" criterion for P T once we publish it, not just for page views. Wikipedia is itself a delicious target for spin. And as artificial intelligence (and other techniques such as vanilla search engine optimisation but AI is the biggie) is increasingly used to promote products and other agendas on the Internet and elsewhere, this will become even more critical if we are to protect our NPOV. And I don't think WP:BALL prohibits us from discussing such issues, nor from taking proactive action if we have good reasons to fear such events. Andrewa (talk) 02:03, 16 July 2019 (UTC)
  • Andrewa, would you please explain why you conclude "there seems a high likelihood that this is just a temporary Primary Topic" (presumably you believe it being temporary is significantly more likely than the average Primary Topic article is likely to be temporary), or, even if so, why this should override WP:CRYSTAL? Thanks! --В²C 19:35, 15 July 2019 (UTC)
    • Agree that WP:CRYSTAL complicates this, and I already pointed that out. [5] Andrewa (talk) 22:35, 15 July 2019 (UTC)
      • Yes, but you have not explained how to address [[WP:BALL], and you have not provided basis for the claim that this is highly likely to not remain primary. --В²C 22:46, 15 July 2019 (UTC)
        • Give me a chance! Working on it. Andrewa (talk) 22:55, 15 July 2019 (UTC)
    • WP:CRYSTAL starts off Wikipedia is not a collection of unverifiable speculation or presumptions. Wikipedia does not predict the future. All articles about anticipated events must be verifiable, and the subject matter must be of sufficiently wide interest that it would merit an article if the event had already occurred... (my emphasis). This isn't nearly as sweeping as is often assumed. It obviously doesn't forbid us from predicting, on the basis of current page stats, what next weeks' page stats will be, and if it did page stats would be irrelevant. And the question here is similar. Do we have good reasons for expecting that the page views will drop when the show ceases to be as newsworthy as it is now? I think we do (and I'm not alone in this), and as we do, it's legitimate to take this into account when deciding on an article title. Andrewa (talk) 02:03, 16 July 2019 (UTC)
      • Andrewa, I too have no doubt the page views will drop when the show ceases to be as newsworthy as it is now. Who wouldn't? The question is how much will they drop, and most importantly, how low will they go relative to the other uses of the term "the americans", which are all very obscure? The show has already been over for several years, and has not really been in the news much at all, but the page views continue to dominate. What good reasons are there to believe they will drop much more than, say, the page views for Law & Order, which is off the air for almost ten years, or L.A. Law which is gone over 20 years, and still gets 10x the views of The Americans (photography)? [6]. Now, if in 5, 10, 20 or 40 years the show does become that obscure, which I admit is a possibility, though don't agree there's a high probability, we can move it then, as the closer noted. But why penalize all the users searching for the show with "the americans" in the mean time? What good reasons are there for that? --В²C 17:24, 16 July 2019 (UTC)
        • Agree that it's important to look at the size of the probable drop. Not only its most likely size, but the range of credible sizes. The whole distribution. And again, I don't think WP:BALL prohibits this, do you?
        • But there is no proposal to penalize anyone, and never has been. If the article had been left at its unambiguous name, all that these people needed to do was to follow an obvious link on a short and quick-loading page, an activity that takes less time to do than to describe here and requires only skills that any one of my four nieces had mastered by the age of six. And any browsing platform should make this just as easy as Explorer and Chrome and Opera do on a PC. The proposal was simply to keep the more logical article name, in order to avoid any problems. Andrewa (talk) 18:31, 16 July 2019 (UTC)
          • If landing on a dab page instead of on the sought article was not considered a "penalty" (relatively speaking), then there would be no primary topics; not even Paris, Blue or God. That ship has sailed. Even your proposal acknowledges that. So there is a penalty to landing on the dab page for "The Americans" rather than on the article about the series for the vast majority searching with "the americans" because they are looking for the article about the show. If you refuse to recognize that penalty, then that explains everything. --В²C 18:39, 16 July 2019 (UTC)
            • I could similarly say, the problem of loading a DAB that clearly links to the page the reader wants is negligible. If you refuse to recognise that reality,.... But such emotive language isn't really helping, any more than talking of a proposed penalty for readers is. There is no such proposal. They have done no wrong!
            • But I acknowledge that it is difficult, in that you and others (myself included) have put many hours of hard work and sometimes passionate discussion into getting Primary Topics right. In your case, particularly in promoting objective measures such as page views, and trying to get away from the more subjective ones such as long term significance. To throw much of that away requires a great deal of courage and serendipity (hmm, that article might be another good test case... it should IMO stay right where it is).
            • Agree that That ship has sailed. Even your proposal acknowledges that. But again the term acknowledges leaves a lot to be desired. My proposal not only acknowledges that, it leverages that. We have this asset and should use it. It might have been an advantage to have deprecated Primary Topic right from day one. But that's academic, not worth discussing, and I thought the proposal already made this clear. Andrewa (talk) 23:38, 16 July 2019 (UTC)
          • Also, I agree it doesn't take a crystal ball to predict the likely foreseeable future of page views based on recent page views. What does take a crystal ball is to predict an eventual relative downturn relative to other uses soon enough that primary topic treatment is not justified. --В²C 20:42, 16 July 2019 (UTC)
            • Progress. The question is, again, where to set the bar? And in this case the bar is I think quite clearly too low, that's one reason I propose it as a test case. But if you're right and page views don't drop sufficiently by the time the RfC is finally moved, then perhaps it won't prove anything. On the other hand, as soon as they do drop sufficiently, it will then become an excellent example of a P T blunder, IMO. Andrewa (talk) 23:38, 16 July 2019 (UTC)
              • No! If you bet that I will not get a six with one roll of a fair die, and I do roll a six, would that be a blunder? No; you bet correctly (you had a 5/6 chance of winning), you just had bad luck. That is, if you were do it over again, you'd bet the same (assuming you're rational and the die roll remains random). Same with PTs that change - just because page view counts sometimes change enough to change PT results does not mean those articles should not have been PTs in the first place. We use the criteria we have and make the best decisions we can with the information that we have. If that information changes in some cases in the future that does not mean we blundered previously in those cases. So no matter what the page view count situation for "The Americans" is in the future, we recently made the right PT decision. It was no blunder, no matter what happens in the future. When only with hindsight can you know whether a given decision was a blunder, then it really wasn't a blunder. To be a true blunder, the information that it is a blunder has to be available at the time the decision is made. You had no way of knowing I'd roll a six, so your bet that I wouldn't wasn't a blunder, even though I did roll a six. We have no way of knowing whether the page views for "The Americans" TV show will drop sufficiently relative to the other uses, or that some other use will arise, to make the show no longer PT, so the decision to treat it as PT now was no blunder either. --В²C 00:29, 17 July 2019 (UTC)
                • That is such strange reasoning that for the moment I'm not going to analyse it any further than that.
                • If the article now at the base name proves to be Primary Topic only temporarily, then IMO moving it there was a blunder. You disagree? Andrewa (talk) 04:35, 17 July 2019 (UTC)
              • Agree The problem of loading a DAB is negligible. The problem isn't the loading of it. The problem is that the DAB is a page that is not the sought article; the reader has to read something else (and some DAB pages can be a real multi-page pain with desired links under less than intuitive headings, for example) before finding the link to the desired page, clicking on it, and finally getting to the sought article. While I was binge-watching The Americans I would look it up every night. And going through that dab page every time was a drag. It was a penalty, a penalty I did not deserve, for, indeed, I had done nothing wrong. And, yet, I was being penalized. Thankfully, we have rectified the situation. --В²C 00:37, 17 July 2019 (UTC)
                • This generic problem would be much better fixed if every DAB page were suffixed with "(disambiguation)". --SmokeyJoe (talk) 01:36, 17 July 2019 (UTC)
                  • Agree that would be a very good idea. But it doesn't solve all the problems. Andrewa (talk) 04:35, 17 July 2019 (UTC)
                • Penalty means punishment. Nobody was punishing you... except perhaps yourself by not bookmarking the page in some way. Andrewa (talk) 04:35, 17 July 2019 (UTC)
                  • If it was some obscure page then it would be reasonable to get me there through multiple links via a dab page and to expect me to create a bookmark for it. Yes adding “(disambiguation)” to the DAB title solved the problem because with the DAB at The Americans (disambiguation) the primary topic article could be at The Americans, which solved the problem. In this case “punish” or “penalize” means to treat unfairly harshly. —В²C 06:34, 17 July 2019 (UTC)
                    1. In the interests of communication, I will try to adopt your understanding of terms such as penalty. I think they are rhetoric and unhelpful, but I must also say that I do not intend any of this as criticism of your behaviour here, just your logic. I am finding this helpful overall in clarifying my own thoughts, and hope you do too.
                    2. Whether the page is obscure has nothing to do with it. The question is rather, how does it stack up against other pages? And it doesn't. It is the flavour of the month, but there is no evidence that it will continue to be so. If it doesn't then it will become a poster boy blunder (like the failure of the earlier New York RMs), which is why I consider it a good test case for the moment. If it does there are still issues but it's not such a good test case, and yes, there have been cases where similar media events have attracted an ongoing cult following, most don't but your obsession with this one does indicate that it might.
                    3. Yes, it's reasonable to get you there through one (not multiple) extra, easily navigated link, if that saves others loading large pages. It's reasonable to expect that if you find this such an unfair and harsh penalty, you would by the third or fourth experience of it avoid any need to go through the DAB by creating some sort of bookmark. The bookmark would also save you from the need to use the search engine. You found repeatedly going through the search engine no penalty apparently? Wasn't it a longer and slower page than the DAB? How many keystrokes did it take to do the search? To fail to do so is irrational, and we can't protect you from that.
                    4. On this occasion the wrong article with which you wish to punish those who adopt rational search strategies is short. But in other cases it will be longer, and our procedures and policies need to deal with these too. There are other cases (and again, New York is a poster boy here) where the pages have been long.
                    5. Finally, no, adding (disambiguation) to the DAB page title doesn't make the base name available for an article. The rules for the destination of the base name are (and should be) unaffected, just as the availability of the name New York City for the city article didn't necessarily make the base name New York available for the state article. This is a common misconception and perhaps a subtle point, and so deserves its own section, see #Name availability. Andrewa (talk) 10:03, 22 July 2019 (UTC) I numbered your bullets for ease of reference. --В²C 01:01, 23 July 2019 (UTC)
  1. Yes, this is helpful for me too. Thank you.
  2. Yes, we agree, the question is, how does it stack up against other pages? We disagree about whether it does. On what grounds do you believe it does not "stack up"? Clearly on page views, it does. Flavor of the month? Seriously? It has dominated like this for years, and there's no good reason to believe that it won't continue to do so. Above, I've provided examples of much older comparable TV shows that remain comparably likely to be sought decades after they went off the air. Even if page views drop off sufficiently so that it no longer meets the usage criteria, it won't have been a blunder because it will have been the right decision at the time it was made. Whether a user chooses to use a book mark is besides the point. Our job is to make article finding as swift as we reasonably can for as many users as we can. Identifying primary topics and choosing titles accordingly is key to this. Obviously if the relatively high page views just started and are likely to drop about as quickly, due to RECENTISM, we're not going to move articles around. But this situation is far from RECENTISM.
  3. One extra link is two; two is multiple ("having or involving several parts, elements, or members"). Saving someone who is looking for a relatively obscure use from loading a wrong page should not be prioritized over swiftly sending the majority directly to the page they seek in deciding how we set up our pages to manage disambiguation, especially if the "solution" is to send all users searching with the term in question to the wrong page (the dab page might not be the wrong article because it's not an article but it is the wrong page because it's not the page being sought). Doing so would fly in the face of the main reason for applying disambiguation techniques: enabling swiftness in finding the sought article.
  4. I presume we can agree to not rehash wrong article here. Suffice it to say it's under dispute, at least by me.
  5. Oh, please. You should have known what I meant. In that particular case, adding (disambiguation) to the dab page did make the base name available for the article because that's why it was renamed like that. As you also should have known, I'm well aware of WP:PRIMARYREDIRECT and related concepts, and you should know better than to think I needed to be schooled about that. --В²C 01:01, 23 July 2019 (UTC)
Most of this seems wrong to me. For now can I just clarify one point as it's obvious that there is a misunderstanding somewhere! You say One extra link is two. Two what? If I land on a DAB page that contains a clear link to the article I want, I can then get to that article in one mouse click. But it takes you two? Andrewa (talk) 19:38, 24 July 2019 (UTC)
There is a click on the original search. That either takes you to the desired article, or to a dab page at which there is an extra/second click required. --В²C 19:59, 24 July 2019 (UTC)
OK, and I've just achieved a PB as that resolves the silliest discussion in which I have yet been involved on or off Wikipedia. When you said via multiple mouse clicks you meant via two mouse clicks and that included the one that you would have needed anyway. I thought I had been clear when I spoke of one extra mouse click, and you never disputed this. We agree that there is only one extra mouse click. Your describing that as multiple mouse clicks misled me, and I take some responsibility for that. Andrewa (talk) 20:27, 24 July 2019 (UTC)

───────────────────────── To return to the point, getting to a DAB that clearly links to the desired article (in one mouse click) seems an insignificant problem to me, to the point that I'd call it no problem at all. On the other hand getting to the wrong article is sometimes a significant problem.

We must accept that one user finds getting to a DAB unacceptable. But I'm skeptical that this is reason enough to move the article and DAB as was done, in terms of overall reader experience. However it was apparently justified by our current policies and procedures, and so it took place, and I think that was bad, and that's the point of this particular test case. Andrewa (talk) 20:36, 24 July 2019 (UTC)

We agree it was extraordinarily silly, but I'm glad we have an understanding on at least that insignificant point now! I wouldn't say landing on a DAB is "unacceptable". I would say it's comparable, and often a bit worse, than landing on the wrong article which has a hatnote link to the correct article. What's unacceptable is when the majority of those searching with a given term have to unnecessarily pay the price of not being taken directly to the article they seek. If it's a minority that are not being taken directly to the sought article - just those searching for a non-primary topic landing on a DAB or on a wrong article with a hatnote link to the sought article or to the DAB - that's acceptable. This is why the usage criteria is worded how it is: highly likely—much more likely than any other single topic, and more likely than all the other topics combined—to be the topic sought when a reader searches for that term. --В²C 20:52, 24 July 2019 (UTC)
Yes, and you want to eliminate the significance criterion and just have page views as the only test, correct? Commendable as that is in some ways, in many cases it just doesn't work. But we agree that Primary Topic as it is could be improved, we just want to move in opposite directions.
landing on a DAB is ... comparable, and often a bit worse, than landing on the wrong article which has a hatnote link to the correct article. Consider;
  • Even articles with a hatnote may have one to a DAB rather than directly to other topics. As you find the one extra mouse click of landing on a DAB rather than the sought article a penalty, I guess this extra mouse click (which would have been avoided had the reader been taken straight to the DAB) is also a penalty, imposed this time on all who don't seek the Primary Topic?
  • At the risk of being repetitive, it's hard to see how loading a DAB can be comparable, and often a bit worse, than loading a long article. Surely loading a long article is a greater penalty than loading a DAB?
You've expressed this opinion many times. It supports your passion for page stats, but there seems no other reason to think it's true. Andrewa (talk) 10:37, 5 August 2019 (UTC)

Name availabilityEdit

From above Yes adding “(disambiguation)” to the DAB title solved the problem because with the DAB at The Americans (disambiguation) the primary topic article could be at The Americans, which solved the problem. [7]

If indeed it solved the problem it certainly wasn't because with the DAB at The Americans (disambiguation) the primary topic article could be at The Americans. But that's a common misconception. It was similarly argued that because the name New York City was available for the city article, the name New York was available for the state. But it's not. In both cases, if there's a Primary Topic, then that should be the destination of the base name, and if there's no Primary Topic then the destination should be a DAB. That's the policy. Andrewa (talk) 10:03, 22 July 2019 (UTC)

No, that's not the same argument. In this case I was talking about the simple obvious fact that if the dab page is at the base name, then the article could not be at the base name. Therefore if the dab page was moved, that reason to not put the article there no longer applied, and, if the topic was primary, then its article could be moved to the base name, or the base name could redirect to the article. Otherwise, if there was no primary topic, then the dab page should be moved back to the base name, or there could be a an odd redirect at The Americans to The Americans (disambiguation). But that's nothing like the New York argument which was that the city did not have a primary topic claim on "New York" since it was at "New York City", flying in the face of WP:PRIMARYREDIRECT. I'm surprised you're conflating the two arguments. --В²C 01:10, 23 July 2019 (UTC)
Again, let's try to get away from the rhetoric... whether you're surprised is not relevant to the logic, and whether or not it's conflating the arguments is the very issue at hand.
The arguments are not identical, but they contain an identical inference... that if an article name is not being used for another topic, then that makes it available regardless of Primary Topic. That's what was asserted in NYRM, and it seems to me that you also appealed to that principle above.
And it's not true. An ambiguous name is only available for an article if the topic of the article is the Primary Topic of the name. Are we agreed on that much?
If so then we can look at why I think it applies to The Americans. And if not, I'll have learned something! Andrewa (talk) 23:40, 23 July 2019 (UTC)
NO!!! I made no such inference or appeal to that silly neophyte notion. YES of course we agree that an ambiguous name is only available for an article (as title or redirect to it) if its topic is primary for that name. And that was the case for The Americans... once the dab was moved out of the way... Reread my original statement. I even explicitly identified it to be the primary topic article. Nothing inferred there. Definitely not “regardless of PT”. —В²C 06:22, 24 July 2019 (UTC)
I think I have again been misled by your rhetoric. I find it unhelpful. For example you now call the mistake under discussion a silly neophyte notion, with which I agree, but on the other hand it is one that was seriously proposed at NYRM and seemed accepted there by some highly experienced editors.
To the point at hand. You said ''Yes adding “(disambiguation)” to the DAB title solved the problem because with the DAB at The Americans (disambiguation) the primary topic article could be at The Americans, which solved the problem. (my emphasis) What exactly was the problem that adding “(disambiguation)” to the DAB title solved? Andrewa (talk) 19:51, 24 July 2019 (UTC)
First, a "highly experienced editor" is not necessarily an expert in the area of title decision-making. But title experts like you and me are knowledgeable about this stuff; other "highly experienced editors" are not necessarily. Second, the problem in this case was that the dab page was at the base name. It had to be moved to the (disambiguation) title before the primary topic article could be moved to the base name. So adding "(disambiguation)" to the DAB title solved the problem. Are you getting enough sleep lately? --В²C 16:43, 25 July 2019 (UTC)
My definition of expert is someone who has already made most of their mistakes. That's how they get to be an expert. Beware the person who thinks they have never made a mistake, they know nothing. Be even more careful if they think they never will, they will never know anything. Or as my essay How not to rant puts it, In the extreme that you never change your mind, you remain an intellectual infant.
I'm very glad you think we are both experts! But my point was not that a highly experienced editor is an expert, just that a mistake made by such editors is not a silly neophyte notion. That's more unhelpful rhetoric IMO.
To call that the dab page was at the base name a problem begs the question. It was only a problem if something else belonged there, in which case there was no problem moving it, as was done. The only problem was if these moves took place and the something else didn't belong there, as I believe to be the case. But I could be wrong. We must wait and see. Andrewa (talk) 09:30, 27 July 2019 (UTC)

How the dropdown works NOTEdit

See for a screenshot.

Note two absences from the list. The Americans (2013 TV Series) isn't there, and neither is The Americans (disambiguation).

There is a link to The Americans, the article that we have decided most readers want right now, and which is about the 2013 series, and to individual articles on the six series, and to The Americans (1961 TV series), and to two unrelated articles.

But what is blatantly missing is any unambiguous reference to The Americans (2013 TV Series). The detailed articles could be about it (and in fact are) but by their names they could also be about The Americans (1961 TV series). And as, going by the list, Wikipedia does cover the 1961 series but not the 2013 series, it might even be logical to (wrongly) expect them to be about the 1961 series.

Wouldn't it be far more helpful to have an unambiguous link to the 2013 series? Andrewa (talk) 22:33, 5 August 2019 (UTC)

At best it would be slightly more helpful, and only for those with the searchbox dropdown; for everyone else it would be far less helpful. The whole point of primary topic is almost anyone searching for a primary topic like "The Americans" would be looking for the series, and would know the unadorned one displayed must be the series. And even if someone doesn't know, seeing the list of particular seasons of the series strongly suggests the first one is the series itself, and even if they don't know and just click on it, odds are they will get to the page they're seeking; that's the whole point of primary topic. Most importantly PT was conjured back before there was no searchbox drop down, and is intended to work for people who are not using it, like anyone who has javascript turned off. We can't decide our titles based on the assumption that everyone is getting the searchbox dropdown list; in fact, we have to assume they don't. --В²C 23:57, 5 August 2019 (UTC)
for everyone else it would be less helpful... Exactly what is the damage of having an unambiguous name on the dropdown list?
would know the unadorned one displayed must be the series... Completely unwarranted assumption, as your later discussion admits. How do they know we even have an article on it? They and you are both guessing. If we have an unambiguous name on the list, no guessing is needed, which is good.
We can't decide our titles based on the assumption that everyone is getting the searchbox dropdown list... Agree. Straw man. Nobody is suggesting we do that.
in fact, we have to assume they don't. Rubbish. We should assume some do and some don't. But perhaps that's what you mean. In which case, we should cater for both those who do and those who don't. That's not assuming they don't. Is it? Andrewa (talk) 03:16, 6 August 2019 (UTC)

And how it now works for NYEdit

By way of contrast with The Americans, where there's an article at the ambiguous name, the dropdown box works rather well for New York, where there's a DAB.

Have a look at where you will see two screenshots.

The first is what you get when you start to type New York. It's a very helpful list, as it has unambiguous links to every article you might possibly want, but not to the DAB. As has been pointed out, our target reader wants an article, not a DAB. It's a definite win.

But where it gets really good is when you get to the screenshot below it, which is what you get when you complete typing New York. There's now an exact match for what you've typed, and almost as if it could think, the computer has now added the DAB to the list.

That's both rather cute and very functional (a rarity in coding, normally cute is to be avoided).

Why does it work so well? Because both the coding and the page names make sense. New York is ambiguous, so we have a DAB there, and naturally disambiguate New York City.

But if we were to decide that New York City was the Primary Topic of New York (as it probably is), and move the article and DAB accordingly, it wouldn't work nearly so well.

Food for thought? Andrewa (talk) 04:53, 9 August 2019 (UTC)

It is already a problemEdit

I'm cross-posting this reply here as, on reflection, it already shows an important point. The question asked was Why would they have to agree with us on PT? [8]

Suppose that a reader thinks that The Americans means the earlier series (ie in their opinion that's the Primary Topic), but they want the later one, and they use the search box. The dropdown list they currently see has no entry that they think means the series they want. The information they want is there, but we have hidden it away under titles that they assume mean something else... most notably, under the ambiguous title The Americans, which they think means the earlier series, but we think means the later one.

If the article they want were at an unambiguous (and dare I say therefore more recognisable) name there would be no problem.

We can of course tweak our search box to fix this. But the more logical fix is, just use unambiguous article names. Anything else is just asking for trouble. Andrewa (talk) 19:20, 7 August 2019 (UTC)

First, if this is truly a problem, it was one introduced with the search box drop down list, so, yeah, that's where it should be fixed, not by redoing our whole system of article arrangement.
Second, if most people are searching for a particular topic associated with a given name (which is the case when consensus agrees there is a primary topic for that name), then most people will also expect that topic to be there. So the users who expect some other topic to be the primary topic and presume the unadorned choice in the drop down list is not the one they are looking for (but it is) have to be a small minority.
Third, the rare reader this affects is most likely going to select the base name anyway, expecting it to take them to what they think is the primary topic article with a hatnote to their article or at least to the dab page. Imagine the pleasant surprise when they arrive at their desired article. What's the problem, really? --В²C 22:29, 7 August 2019 (UTC)
First point: Do you really think there is no problem? Wouldn't it be better to have an unambiguous link to the 2013 series? It would be. Get over it.
In reply to this point, yes and no. It could be fixed by tweaking our software to cover for our illogical naming of the articles, yes. But that's not the best fix.
Second point: Disagree. Baseless speculation, and misrepresentation of what P T currently means (that's what you'd like it to mean, yes, but there's no consensus that it's desirable or even possible). Not everyone thinks and works the way you do. Again, get over it.
Third point: Passing over the speculation that this reader is rare and the idea that if they are rare then they don't matter, imagine the disappointment when, having loaded a long article on a slow connection or a mobile device, they find it's not the right article anyway. That's not a problem? It is.
The logical way to fix all of the problems is to heed the advice of the very first few words of the article title policy: Article titles should be recognizable. An ambiguous title is less recognisable, both by the reader and by the software (assuming the software is logically written). So they should be avoided wherever possible. Andrewa (talk) 22:11, 8 August 2019 (UTC)
I think the problem, if any, is minor and isolated to a very strange thing - the search box drop down list - the use of which is optional; totally unnecessary. What a snoozefest. We do have an unambiguous link to the 2013 series. It's The Americans (2013 TV series) which is an unambiguous link to the 2013 TV Series at The Americans.
The community doesn't always agree on whether there is a primary topic or on what it is, but generally it makes sense that what we think is the primary topic is what most readers would expect to find at the base name, or at least wouldn't be surprised to find it there.
It's 2019. Who even has a connection that slow any more? Certainly fewer than did back when WP was launched. And we had primary topics then. Get over it.IWe can't reorganize our articles for all these tiny special cases. --В²C 22:41, 8 August 2019 (UTC)
Again, nobody is proposing anything based on all these tiny special cases. More rhetoric. It's helpful to consider actual examples. You'd prefer not to?
Yes, We do have an unambiguous link to the 2013 series. It's The Americans (2013 TV series) which is an unambiguous link to the 2013 TV Series at The Americans, and that's good. But we only have it because the article was once there, and that's bad. We should have unambiguous links to all articles.
A workaround to that problem is suggested in the proposal here. But much better to fix the logic of our article naming policy instead. Andrewa (talk) 23:42, 8 August 2019 (UTC)
Oh yeah, having that link is SO important[9] (rolls eyes). Dude, you need to find something with more impact to do with your life. --В²C 01:02, 9 August 2019 (UTC)
A spectacular result indeed... no traffic at all via the redirect! Or that would be the naive interpretation. But I had myself used the link in that period, just to test it, so I found it a bit hard to believe. Didn't you?
But this on the other hand gives a rather different picture. So, how do you interpret your chart now? Andrewa (talk) 02:24, 9 August 2019 (UTC)
Sigh. The links to it, of course. —В²C 06:20, 9 August 2019 (UTC)
So, when you said Oh yeah, having that link is SO important [10] (rolls eyes), you knew that hundreds of users per day were using that redirect (link as you call it), is that correct? Andrewa (talk) 15:19, 9 August 2019 (UTC)
I had noticed that recently but I had forgotten about it at the time I wrote that statement and then was reminded about it after your reply prompted me to look at it again. Such artifact links are typical after any page move (regardless of whether the move is due to PT) and they're benign. But the point is if not for this artifacts from the article being at that title previously, there would be little if any benefit from that redirect. It's not like they're knowingly "using" it - if all those links were fixed to go directly to The Americans they wouldn't know any different. --В²C 18:20, 9 August 2019 (UTC)
I'll take that as a no. The link I gave made your rolls eyes comment ridiculous.
In fact, you were badly misinterpretting the data you presented. Each day there were hundreds of people using the link that you said didn't matter, and you were aware of that possibility, but overlooked it when you wrote what you did. And anyone naively looking at that comment and the chart you presented in good faith but in error would have been misled. Is that an unfair assessment?
And that's the problem with your recommended approach to page stats and Primary Topic in general. All statistics need to be interpreted. Some are trickier than others. The page stats tools are a nightmare, in that they can look so very convincing in specific cases, but then on closer examination (if such is allowed!) we form consensus that they are misleading.
Most of the rest of that post seems to just avoid the question. Andrewa (talk) 23:15, 9 August 2019 (UTC)
You're making a mountain out of less than a molehill; out of nothing. To what end? The problem in my link was using upper instead of lower case s for series/Series. This link is correct, showing when the title change occurs. The counts of the disambiguated previous title (now a redirect) include "views" that are due to links to the article through this redirect. If those are all corrected, the links will go direct to the article at The Americans, and so will the page views. In any case it demonstrates that it is unimportant to our readers beyond the temporary technical function it must fulfill after a title change until the old links are fixed. None of this supports your claim: "We should have unambiguous links to all articles", which precipitated this current thread. --В²C 23:56, 9 August 2019 (UTC)
OK, so you've finally found out why the table that you so confidently cited was a load of rubbish. A simple typo.
But it's so easy to do, as we have just demonstrated. And you claim to be an expert on the use of page stats I think, and I haven't disputed this, and I'm certainly not a page stats expert. But I do have enough training in Mathematical Statistics to often scratch my head at your approach. One thing you should ask of any such result is, is it reasonable? That's one thing the current Primary Topic criteria seek to do, but you not only fail to do so, you want the rest of us to stop this reasonableness or significance checking too.
Just the same I don't dispute your idea that to go by page stats alone might be better than what we are doing now. It would fix some of the problems, but introduce others, and I still think there's a better way.
So far as making a mountain out of a molehill or less is concerned, I think that may be the pot calling the kettle black.
I don't think the rest of the post even addresses the question of whether we should have unambiguous links to all articles. We've demonstrated that it helps to have them. You seem to think that it does a little damage, but have given no reason for this belief.
And nor does it or anything else you have posted address the scenario with which I started this whole section. [11] You've said that it's rare, but that seems to be just a guess. Whether this user is disadvantaged by the ambiguous name is my first question. I think they are. You don't? Andrewa (talk) 04:26, 10 August 2019 (UTC)

It is still a problemEdit

This edit raises some interesting points. Watch this space. Andrewa (talk) 02:36, 21 August 2019 (UTC)

Actively damaging the productEdit

The Americans (2013 TV series). I was good to ignore the The Americans (2013 TV series) titling dispute as a very unimportant albeit popular topic. A commercial product, popular, and associated with tremendous click bias due to the sort of people interested, regarding their ease of access. Pageviews are maximised by idea people jumping from one transient short term interest to another. It barely fits the goal of the project, but at least it doesn't do actual harm.

Or does it?

Watching В²C for his next random project damaging exercise, I see he is on a spree of removing useful hovertext. It is in violation of WP:NOTBROKEN. It does nothing for content, the content is unchanged by the edit, the only effect is the removal of hovertext. The purpose of hovertext is to provide clarifying information for people actively seeking that information. B2C is actively removing it. Why? --SmokeyJoe (talk) 05:44, 20 August 2019 (UTC),_Elizabeth&diff=prev&oldid=911635938

My understanding is this kind of link cleanup is not required, but not a problem either, and arguably desirable to get rid of piped links. In this case I don’t think the hover text is useful. If I find an article where it is helpful, I’ll leave it. I’ve brought up this effort generally and specifically in two places. See:
В²C 07:12, 20 August 2019 (UTC)
    • PRECISE Hovertext is always useful, especially to a highly ambiguous title for readers not familiar with DIFFCAPS, and it never hurts. It is true that counting incoming links is easier if piped links to redirects are removed, but this redirect is the full PRECISE title, and thus is special, worth leaving along even if it does mean an extra step for the hypothetical incoming link counter. --SmokeyJoe (talk) 07:41, 20 August 2019 (UTC)


There seems no harm in leaving Mathematics at the base name, and no chance of consensus for moving it. It's not even obvious where we would move it.

The hatnotes to mathematics (disambiguation) and math (disambiguation) are effective navigation devices, and the article is relatively short.

Perhaps, there should be a policy that all articles at ambiguous base should be short? Andrewa (talk) 21:48, 6 July 2019 (UTC)

I've added that as a possible workaround, see User:Andrewa/Primary Topic RfC#Alternatives. Andrewa (talk) 19:15, 10 July 2019 (UTC)

New YorkEdit

This is a case of Primary Topic proving ineffective.

Various RMs over a period of more than ten years failed to dislodge the New York (state) article from the base name New York, despite clear consensus that it never belonged there under any policy, and particularly that it was not the Primary Topic.

Many thousands of mislinkings were created within Wikipedia during this time, most of them by linking to New York assuming that this would link to the city. These have been fixed. Had the New York name been a DAB, the creators of these mislinkings would have received an automated warning, and even if they had ignored this Wikignomes would have been easily able to find and fix the mislinkings.

It's reasonable to conclude that there remain many thousands of incoming external links pointing to the wrong article. We have no way of fixing or even detecting these, and never will have.

There is a strong case that the Primary Topic of New York is New York City. It is both the more likely search term and the more significant topic. However, it is better for the New York City article to remain at an unambiguous name, and for the destination of New York to remain a DAB, for many reasons. Andrewa (talk) 22:11, 6 July 2019 (UTC)

From User:Andrewa/P T test cases#New York: The rare if not unique problem with New York is that even though the city article clearly gets more page views than the state article does, we have no of way knowing how often people use just "new york" to search for the city vs the state. Is it half of the searches? 25%? 90%? Who knows? I think it's a special case and consensus finally got it right to put the dab page at the base name. If for no other reason other than it's not clear if New York City is the primary topic for "New York". I don't think this is a good example for anything else. --В²C 00:32, 25 July 2019 (UTC) [12]

Why is that any different from say wave, currently an article on the topic of wave (physics)? How do we tell how often people use just wave to search for wind wave? Andrewa (talk) 05:43, 25 July 2019 (UTC)

Independent topicsEdit

SmokeyJoe said Primary Topic is not well defined for New York City vs New York State because they are not independent topics. [13]

That's extremely relevant if true. But is it? Discussion welcome here. Andrewa (talk) 20:24, 26 July 2019 (UTC)

  • PrimaryTopic is a test that suits competing topics, not different aspects of the same topic. NYState and NYC both claim identity-descent from the colony called New York. That makes it complicated with regards to long term significance, much of their long term significance overlaps. In terms of recognizability (like what some clumsily call the likelihood of what someone is searching for), the distinction between city and state is blurred or unimportant. —SmokeyJoe (talk) 23:33, 29 July 2019 (UTC)
    • I think I see what you mean, and if so I agree.
    • But that doesn't save our current policy. The eleven year series of NYRMs was a fiasco of epic proportions, the damage was considerable, and even more important, the good end result is probably not even supported by the current policy. So it's still a good example of how NOT to do things.
    • Perhaps, an even better solution would be an overview article or BCI at the base name? I'm not averse to that at all, I recently saved what is now User:Andrewa/New York (overview) from deletion as an abandoned Draft:New York (overview). Andrewa (talk) 03:51, 30 July 2019 (UTC)
      • PrimaryTopic "policy" (guidance) is often OK, often so-so, and often unhelpful. However, it is not the root of all evils, NY is a difficult case beyond the PT perspective. --SmokeyJoe (talk) 04:15, 30 July 2019 (UTC)
        • Agree it's not the root of all evils. (And neither is money, it's the love of money that is described as this in the Bible. Perhaps similarly, the problems with Primary Topic are principally when it becomes the focus of obsessive behaviour. Interesting thought? I hope I'm not guilty of that, but it's been mentioned at least twice recently, including once by yourself I think, in connection with one particular user.)
        • It's policy to have Primary Topics, see wp:P T. IMO this is usually OK, never helpful (in that there's no problem it solves), and sometimes problematic.
        • New York is an example of a problematic case. It's an extreme, but that doesn't disqualify it as an example.
        • What would be really helpful is an example of a case where the current policy is more helpful than my proposal would be. The only candidate so far is The Americans, which will admittedly rescue people who can't use bookmarks, and who find it easier to click on a hatnote that isn't there than on a link from a DAB that is, and who find it far more (or at least equally, they've expressed both views rather strongly) annoying to load a short DAB (which they insist on calling a wrong non-article page) than a long wrong article (which they similarly insist on calling a wrong article page), and who therefore almost throw their expensive tablets across the room (I hope their mummies know about Gorilla Glass)... Ummm, do we have a better example? Andrewa (talk) 09:51, 30 July 2019 (UTC)

Identity-descent [14] is a good term for it. Andrewa (talk) 10:50, 30 July 2019 (UTC)

Examples where P T is helpfulEdit

So far, all the examples here have been ones I have suggested of cases in which the current policy and guidelines have failed.

There has been some discussion of some of these, and I thank all those who have contributed. But my opinion remains that the current P T policy and guideline combination is usually OK, never helpful (in that there's no problem it solves), and sometimes problematic. [15]

This in response to the opinion that it's often OK, often so-so, and often unhelpful. [16] Read the whole post and make sure I'm not misquoting it, that's what the diff link is for. It doesn't seem too far from my opinion. There does seem room for improvement.

But I'm eager to change my mind if there is evidence to the contrary. Have we any, possibly apart from The Americans? So far as The Americans goes, another editor has commented that having the DAB at the base name was probably best. So it's not just me. The move was decided along lines of current policy/guidelines, not reader benefit.

So I'm after any examples where the reader is better off under the current policy/guideline than under my current proposal. I'm eager to hear (and possibly discuss) suggestions.

I've put a heads-up to this challenge here but this is currently a busier page anyway.

One place to look is at User talk:Andrewa/P T examples and scenarios where there are many more examples collected over the past year, mainly if not exclusively from WP:RM. IMO none of them show any improvement of Wikipedia produced by the Primary Topic policies/guidelines. And of course the ones here at User:andrewa/P T test cases tend to show damage to Wikipedia, again IMO produced by Primary Topic.

So, prove me wrong. Or at least, please have a go. Andrewa (talk) 11:28, 31 July 2019 (UTC)

  • If there were no Primarytopic exception to WP:PRECISE, there’d be no case to introduce it. —SmokeyJoe (talk) 11:55, 31 July 2019 (UTC)
    • Not sure what you're getting at, SmokeyJoe, and it doesn't seem to answer my question anyway! But to reply, WP:PRECISE redirects to the policy section on Precision and disambiguation, which is definitely on the money. The subsection on Precision starts Usually, titles should unambiguously define the topical scope of the article, but should be no more precise than that. I'd of course like to simply delete that word usually. And the subsection on Disambiguation refers to Primary Topic of course, and the disambiguation guideline, which all currently seems consistent with that. Some of this would of course also need to be changed under my proposal, to say that P T only now applied to grandfathered articles (and it doesn't need to be said, but just as now there may still be the occasional exception, which is why we have IAR). Andrewa (talk) 19:11, 31 July 2019 (UTC)
  • In general, any situation where an article that is the primary topic by usage criteria, and is at the ambiguous base name accordingly (or the base name is primary redirect to the article), that's an improvement to Wikipedia. That includes The Americans as well Paris, Jimmy Carter, Queens, etc. etc. In each of these cases, by definition (because it's the primary topic by usage), when a user searches with the title in question they are taken directly to the article they are most likely seeking, rather than taken to a dab page through which they have to wade. Paris (disambiguation), for example, is several pages long. Do we really want to send searches of "Paris" there? Of course not. The average reader is clearly better off in every single case where the primary topic is primary due to the usage criteria. I will agree that no such case can be made where the topic is not primary by usage. So, for example, having the fruit at Apple does not benefit the average reader; putting the company there, or having Apple redirect to Apple, Inc., would benefit the average reader searching with "apple", since they are most likely seaching for the company based on page views. --В²C 18:03, 31 July 2019 (UTC)
    @Born2cycle: why don't you start a RM at Talk:Apple (disambiguation) to put the DAB at the base name and the fruit at Apple (fruit), for the reasons I explained earlier I would definitely oppose to making the company primary. Crouch, Swale (talk) 18:58, 31 July 2019 (UTC)
    You've expressed these views repeatedly, but still have no examples to back them up (with the possible exception of The Americans). Nobody is proposing to move Paris for example. Why do you persist in bringing it up as if I did? Andrewa (talk) 19:15, 31 July 2019 (UTC)
    Because your argument applies equally to Paris. Just because you've made an explicit exception out of it means just that. But if we ignore the exception, then the essence of your proposal argument indicates the dab page should be at Paris. So I think it's a valid example to consider. --В²C 19:19, 31 July 2019 (UTC)
    Let's get this straight... I'm excluding all existing articles, including Paris but not restricted to it. Exactly which argument applies equally to Paris, in view of that? Andrewa (talk) 19:26, 31 July 2019 (UTC)
    Yes, you're making exceptions of all existing articles. But why? Pure pragmatic-political reasons - you know you'll never get consensus support without that (it's hard enough with it). But there is no theoretical justification for these exceptions that are consistent with the essence of your proposal. For example, just above you reflected this essence when you wrote: "I'd of course like to simply delete that word usually.". That usually, the one you would like to "simply delete", is what justifies Paris being at Paris (and The Americans being at The Americans), etc. --В²C 20:05, 31 July 2019 (UTC)
    Straw man having failed we now have ad hominem. But not even accurate ad hominem at that. No, I've changed my proposal following my second epiphany, and this is already well documented but I think probably a new user subpage is called for. Certainly part of the reason was that the current proposal has more chance of success, but the more important reason is that the current proposal is a far better one. And that's why I think it has more chance of success.
    And that essence you want to talk about is another name for the previous proposal. The straw man is up for another battering! Please, let us move on from it.
    But agree that That usually, the one you would like to "simply delete", is what justifies... etc.. Exactly. This vagueness is an essential part of the current policy/guideline, which would otherwise be inconsistent. But under my (current) proposal, this vagueness is unnecessary. And that's a good thing. Isn't it? Andrewa (talk) 23:30, 31 July 2019 (UTC)
    Strawman? ad hom? Huh? We have some kind of big disconnect (again). Probably my fault, but let's figure this out. Working backwards... since under your current proposal you allow for Paris to remain, how is the usually vagueness unnecessary? My understanding: if you delete usually then Paris is contrary to guidelines which would be contrary to your proposal. So the usually vagueness must be retained under your proposal. No? --В²C 23:46, 31 July 2019 (UTC)
    No, the vagueness disappears along with the word. It becomes Titles should unambiguously define the topical scope of the article, but should be no more precise than that, which is perhaps still not perfect but less vague. We should not make perfect the enemy of good (as someone said during the NYRM saga, I forget who but it's an excellent point). And of course corresponding changes are needed elsewhere, perhaps the word simply was misleading, mea culpa on that. The change to this particular clause is simple, that's all I meant by it. Andrewa (talk) 01:25, 1 August 2019 (UTC)

The silence is deafeningEdit

None of the replies so far seem to answer my request for any examples where the reader is better off under the current policy/guideline than under my current proposal (although B2C did try). But I'm not going to argue from silence that there aren't any! There probably are. And I'd very much like to discuss them, and see whether they really justify the current policy and guidelines. Andrewa (talk) 19:15, 31 July 2019 (UTC)

The silence of B2C on providing examples where the reader is better off under the current policy/guideline than under my current proposal is deafening. Instead of discussing the actual proposal, they now want to discuss a previous version of it, a version that they have called its essence. See above.

But I'm still hoping that they or others will provide some examples as requested, so I think we need a subheading in addition to the outdent above to allow a place for this. Andrewa (talk) 23:30, 31 July 2019 (UTC)

Previous version? The Rationale section of the current proposal provides no rationale for grandfathering primary topics for existing primary topics. Thus, the explicit exceptions aside, per the essence/rationale of the (current) proposal, Paris (along with all other ambiguous titles) should be disambiguated. --В²C 00:15, 1 August 2019 (UTC)
(Sigh) I guess I asked for that. This new subsection was supposed to be about examples where the reader is better off under the current policy/guideline than under my current proposal (emphasis added). I'll try again at #Actual examples where P T is helpful below, in the hope that we can keep it on-topic. My fault here, I raised other issues which I'll answer here.
Perhaps the current proposal could be clearer on this point. But there's no intention to move Paris, and I thought that was already clear. Does this fix it? Andrewa (talk) 01:32, 1 August 2019 (UTC)

Actual examples where P T is helpfulEdit

None of the replies so far seem to answer my request for any examples where the reader is better off under the current policy/guideline than under my current proposal (although B2C did try). But I'm not going to argue from silence that there aren't any! There probably are. And I'd very much like to discuss them, and see whether they really justify the current policy and guidelines. (And I think we've given Paris and The Americans enough discussion elsewhere, which will probably continue.) Andrewa (talk) 00:49, 1 August 2019 (UTC)

  • I'm not sure what exactly you're looking for. Any primary topic article is an example where P T is helpful. You're asking for examples of water that is wet. But here's an explanation for one in particular I just ran into; Alabama (band), per Thryduulf. Although this is about P Ts that are PDABs, note that the explanation of P T helpfulness applies to any P T:
    • If someone is searching for e.g. Alabama (band) they are almost certainly going to be looking for the most famous band of that name and should be taken directly to the article they are looking for. Those relatively few looking for the lessor known band will expect less effort getting there from the hatnote at the bigger band's article than they would have to to pick the listing out of the 11 entries in the relevant one of 7 categories at Alabama (disambiguation). Very large numbers of our readers know how we title articles, and very nearly everybody who is looking for the article about the country music band will also know about the state. As such those people who fall into both groups will naturally search for "Alabama (band)", however it is likely that most of those people do not know that there was a different band with that name so will not know of the need to add additional qualifiers, or indeed what other qualifiers would be necessary - (Alabama band), (1980s band), (band formed 1969), (country band), (country music band) even assuming they know these facts before reading the article. [17]
    --В²C 18:25, 1 August 2019 (UTC)
    note that where I wrote "will expect less effort" I meant "will expend less effort". Thryduulf (talk) 18:53, 1 August 2019 (UTC)
    I'm certainly not looking for examples such as Alabama (band) at this stage, as that article is not at a base name and I don't think there's any suggestion that it should be. Such incomplete disambiguations do raise similar issues, admittedly, and if we can't find any examples of reader benefit from Primary Topic then I'm happy to discuss them. Andrewa (talk) 20:16, 1 August 2019 (UTC)

I think what you are asking for is examples of things that are not currently ambiguous and/or do not currently have a primary topic but where readers in the future will benefit from a primary topic more than a base disambiguation if the situation changes? If so then it's very obviously impossible to give specific examples (if it isn't obvious to you then re-read WP:CRYSTAL) but speaking generally it will benefit readers in exactly the same way and for exactly the same reasons as every example of an existing primary topic does.
If for example someone writes a novel called Westferry Road and a band takes its name from that novel. The band become a huge international success, and off the back of that the novel and the street both become notable enough for articles because of attention from fans of the band. The band would be the topic almost everyone will be looking for, so forcing them to go through a disambiguation page would be unnecessary (and would also require extra work from editors cleaning up links to the disambiguation page intended for the band). Those looking for the book or the street will have the same number of clicks from the hatnotes or at most one extra if there is only a hatnote to the disambig page). Thryduulf (talk) 19:21, 1 August 2019 (UTC)

Thanks, and the issues you raise are worth discussion, and I'll make a place for that, watch this space!
But no, what I'm looking for here is actual examples that exist now and give readers overall a benefit from the primary topic being at the base name, and where we would lose this benefit under my current proposal. The Americans has been suggested and I have valued that discussion but I'm unconvinced. Paris has been suggested but would be grandfathered under my proposal so is unaffected. So, have we any others? Andrewa (talk) 20:16, 1 August 2019 (UTC)
I must be misunderstanding your proposal then, because if all current instances where there is a primary topic will remain grandfathered under your proposal then it is impossible to for there to be a situation where there is benefit from a primary topic now but not after your proposal. Thryduulf (talk) 20:24, 1 August 2019 (UTC)
Good point, I've been getting a bit muddled, haven't I? Let me digest that. How do we balance that with wp:BALL? Which was your point before of course. I'm a bit busy IRL for the next few days, but watch this space. Andrewa (talk) 21:09, 1 August 2019 (UTC)

This is why I've been confused. You ask for examples, so I brought up Paris. Then you said that would be grandfathered. Well, everything would be grandfathered, so we can only be talking hypotheticals here, assuming a given article that is a PT now would be created under your proposal instead as not a P T and discuss how that would be better/worse for users. So I return with Paris and you again reject as grandfathered and, well, here we are. So Alabama (band) is grandfathered too. Are we to imagine a new U.S. band, perhaps named Wyoming? Under both current rules and your rules it would go to Wyoming (band). Fine. Now let's say someone in the UK starts a band named Wyoming and it gets created at Wyoming (UK band). Under your proposed rules the original US band could not be a primary topic so it would be moved to Wyoming (US band). But under the current rules we would consider how well known each band is, page views, awards etc. If the US band achieved the notoriety of, say Alabama (band), and the UK band was relatively obscure, then we would leave the US band at Wyoming (band), per the user-benefit argument that Thryduulf made above for Alabama (band). Does this example count? --В²C 21:57, 1 August 2019 (UTC)

That's another example of incomplete disambiguation. I'm concerned with what goes at base names, rather than how we disambiguate when we do disambiguate. My proposal doesn't change any of that as far as I can see.
Perhaps I should expand the proposal, but my immediate reaction is, it just muddies the waters. Better to get base names right and see where that leads. Andrewa (talk) 23:34, 1 August 2019 (UTC)
But surely you recognize that the explanation Thryduulf gave for how it helped readers for us to recognize P T and treat it accordingly applies equally to base name P Ts as it does to PDAB P Ts? --В²C 23:45, 1 August 2019 (UTC)
No, I think that's just muddying the waters. Thryduulf makes a good point as noted above, but that's not it. Andrewa (talk) 19:58, 4 August 2019 (UTC)
So are you saying that for something that isn't currently ambiguous, e.g. Costa Blanca were to become ambiguous we couldn't have Costa Blanca as primary topic with a hatnote to Costa Blanca (racehorse) but only Costa Blanca as a disambiguation page linking to Costa Blanca, Spain and Costa Blanca (racehorse)? Thryduulf (talk) 00:03, 5 August 2019 (UTC)
No, the proposal is for new pages (or renaming old ones). Existing pages (whenever they were created) remain where they are even if they later become ambiguous, providing they meet the Primary Topic criteria. But if they are subsequently moved for any reason (such as there's no Primary Topic, or that another topic now has the nod), the destination of the now ambiguous term becomes a DAB. Any new primary topic remains at its unambiguous name.
Or that is the intention! How can I make it clearer? Andrewa (talk) 03:57, 5 August 2019 (UTC)
So what is the benefit to anybody, but particularly readers, in having primary topics be at the primary title in most cases but not allowing primary topics to be primary if they were created at a non-primary title after an arbitrary date? Thryduulf (talk) 10:43, 5 August 2019 (UTC)
That's obviously another thing I need to clarify! There is no arbitrary date. The grandfathering process continues indefinitely. How could I make that clearer in the proposal? I thought it was also clear above when I said Existing pages (whenever they were created) remain where they are... but I can now see that the existing was ambiguous.
But I think you may be asking, what's the benefit? The page says:
  • Improve the chance of readers finding the article they want.
  • Reduce the number of mouse clicks required to find the correct article.
  • Reduce the number of article moves.
  • Reduce mislinkings within Wikipedia.
(And there's now a fifth line to the rationale which I didn't think was necessary, as it's covered elsewhere in the proposal and isn't a motivation for change, but another editor wanted it clarified and it does no harm.)
Whether those four would actually be achieved is probably better discussed at User talk:Andrewa/Primary Topic RfC#Rationale. Andrewa (talk) 19:26, 5 August 2019 (UTC)
For sake of argument, let's say your proposal as it stands now attains consensus consensus on January 1, 2030 (to be generous about its chances). Then, let's say that year some guy named Hubert Ramsey bcomes very notable, much more notable and clearly historically significant than the current topic at Hubert Ramsey, a relatively obscure lacrosse player. By the current rules we'd be able to put the much more notable Ramsey at the base name after disambiguating the lacrosse player. Would that be still "allowed"? By my reading it wouldn't. At best we'd disambiguate the lacrosse player and put a dab page at the base name, right? How would that help readers, given that practically everyone searching with this name would be looking for the much more notable Ramsey, not the dab page? --В²C 19:44, 5 August 2019 (UTC)
Good example! No way would the new Primary Topic be moved to the base name! Why on earth would we want to do that? What possible benefit would there be? There's a downside, but the most important thing I think is that there's no upside. The often-assumed benefits are mythical. Realising that was of course my first epiphany.
What we would do is to put a DAB at the base name if we had consensus that the subject of that older article was no longer the primary topic, and disambiguate that older article. This would have several benefits, such as avoiding the situation we now have with The Americans, see #How the dropdown works NOT. Andrewa (talk) 22:57, 5 August 2019 (UTC)

|Why on earth would we not want the primary topic at the base name!? Primary topic means, by definition, that it is the topic most people are looking for, so having the article at the title people look for it at benefits them. It also benefits people creating links to that article, as most of them will go to the correct place whether they know the name is ambiguous or not. It therefore also benefits those people who fix links to disambiguation pages (as there will be fewer of them). Your comments about number of clicks are just plain wrong:

  • Scenario 1: Primary topic at base name, hatnote to disambiguation page only.
  • Scenario 2: Primary topic at base name, hatnotes to secondary topic and disambiguation page.
  • Scenario 3: Disambiguation page at base name

Assume that 1000 people are looking for the primary topic, 200 for the secondary and 100 for all other topics combined.

Number of clicks after reaching Wikipedia
Looking for Scenario 1 Scenario 2 Scenario 3
Primary topic 0 0 1
Secondary topic 2 1 1
Minor topic 2 2 1
TOTAL CLICKS 600 400 1400

Don't forget that on a large dab page it can take some time and mental effort to find the entry you are looking for on the dab page (add another click for everyone going via a dab page organised into sections and your totals would be 900 / 500 / 2600. Now imagine the difference when dealing with the page views popular topics get
I don't know what sort of "epiphany" you have had, but your claim that "The often-assumed benefits [of primary topics at the base name] are mythical." is directly contradictory to my nearly 15 years of experience of benefiting from primary topics being at the base name.Thryduulf (talk) 00:52, 6 August 2019 (UTC)

  • That's the argument for repudiating WP:MALPLACED, and suffixing every disambiguation page with "(disambiguation)", which will solve the unwanted disambiguation page problem. It fits equally with respect for precise titling. Ease of wikilinking is a bunk justification. --SmokeyJoe (talk) 00:57, 6 August 2019 (UTC)
  • What Thryduulf said. SmokeyJoe, what is the argument for repudiating WP:MALPLACED? Thryduulf's argument? How so? It seems agnostic regarding WP:MALPLACED to me. I mean, there's no difference in his table for Scenario 3 whether the dab page is at the base name, or there is a redirect at the base name to the MALPLACED disambiguated dab page. I don't see how suffixing every dab page with "(disambiguation)" changes anything, much less solves anything, assuming the base name still redirects to said (MALPLACED) dab page. I don't think ease of wikilinking is bunk, but it is a lower priority than user experience. --В²C 01:17, 6 August 2019 (UTC)
    • Putting the disambiguation page at the basename is a root cause for people going to the disambiguation page unexpectedly. If the basename redirects to a disambiguation page, it will not appear in the Go box list of suggestions. It will not be reported by external search engines as a viable destination page. Further, bots will still be able to find bad wikilinks to these basenames that are redirects to disambiguation pages and assist in getting them fixed. If no one can simply click on something (other that an explicitly titled disambiguation page) to go to a disambiguation page, most of the wasted click cost will not occur. The way I see it, MALPLACED is the sole reason for the over-exuberance of applying PrimaryTopic status to weak claims. --SmokeyJoe (talk) 01:30, 6 August 2019 (UTC)
      • The Go box list of suggestions, SmokeyJoe? The ones that show up only if JavaScript is enabled? We should not be making title decisions assuming everyone is running with JavaScript. —В²C 04:44, 6 August 2019 (UTC)
      • Not every click to a disambiguation page is wasted - you are assuming that everybody knows the title of the article they want. For example if I'm looking for information about an Irish footballer called Peter Murphy do I want Peter Murphy (footballer, born 1925), Peter Murphy (footballer born, 1980) or Peter Murphy (footballer born, 1990)? The context I've seen his name in doesn't give any hints. As for malpaced that is indeed completely irrelevant - the internal search suggestions on desktop currently prioritise article titles over redirects, but these suggestions are only seen by readers who use the internal search engine on desktop browsers with javascript installed and activated. The exact proportion of readers who use this method is not public (afaik) but it is unlikley to be a majority (I think internal links and links from external search engines together will account for much greater than 50% of views). Thryduulf (talk) 09:12, 6 August 2019 (UTC)
        • Good point that I missed before, and which hadn't occurred to me. That's a scenario in which the reader does (or at least should) want to be taken to the DAB, to learn how many footballers by that name have articles here, and to start the process of deciding which one they want. Andrewa (talk) 06:55, 10 August 2019 (UTC)
  • I also want to remind Thryduulf of the distraction cost of hatnotes. The prime realestate of an article is immediately below the title. The title is what was read in choosing to go to the article. The Hatnote tells the reader things about everything but what the article is about. Every reader who wants the page has to read the hatnote. The hatnote is horrible with screenreaders (I am working with screen readers, they are mostly horrible already granted). The hatnotes display differently in different formats. Hatnotes may be needed, but should be avoided where possible. Precise titling eliminates many of the needs for hatnotes. --SmokeyJoe (talk) 01:36, 6 August 2019 (UTC)
    • Hatnote links his are very useful. I frequently take advantage of them. The tiny minority of users using screen readers should not be causing us to remove very useful features for the majority. —В²C 04:44, 6 August 2019 (UTC)
      • They are very useful. I use them too. They are indeed virtually essential when ambiguous titles are used. Just because they work doesn't mean they are the best solution. Reader accessibility is a real issue with increasing legal obligations. It is not ok to discount accessibility concerns based on low numbers. No one wants to remove a needed useful feature, but if there is another way ... such as a precise title eliminating the use of the hatnote pointing to multiple alternative meanings of the ambiguous title. --SmokeyJoe (talk) 05:26, 6 August 2019 (UTC)
        • The correct solution to the problem of screenreaders not working well with hatnotes is to fix screenreaders and/or inject metadata into hatnote templates to help screenreaders deal with them better. We improve the lives of disabled people by making things better for disabled people (and others as well) not by making things worse for everybody else (which frequently results in things getting worse for disabled people as well). As long as there are titles that are potentially ambiguous (which will happen regardless of how precisely we title articles) hatnotes are indispensable. Thryduulf (talk) 09:12, 6 August 2019 (UTC)
          • We help everyone when article titles are logical. Deliberately using ambiguous terms as article names is just not logical. It is bound to lead to trouble and does. Andrewa (talk) 11:27, 6 August 2019 (UTC)
            • Meanwhile in the real world, making main articles harder to find is not logical in the slightest. We help many orders of magnitude more people when primary topics are at the base name - when the primary topic is at the base name most people get the article they want immediately, when disambiguation is primary nobody does. Do you really think that if the current situation was as bad as you claim that it would have lasted 18+ years? Do you really not recognise that if your proposal was really so much better that there wouldn't be a need to grandfather anything in? Thryduulf (talk) 12:20, 6 August 2019 (UTC)
              • I wonder if the main factor that causes the disparate opinions about primary topic is what one imagines to be the target user relative to a subtle little feature that I called the search box drop down list. If your target users are those searching using that drop-down list, then totally unambiguous titles make sense. If your target users are entering terms without having any feedback hints, perhaps because JavaScript is not enabled, and hitting Search (or especially GO), then ambiguous titles for primary topics make sense. People imagining the dropdown-using type of target user would explain why support for primary topic seems to have lessened since the introduction of this feature. But I think we need to arrange our articles as if that feature does not exist. That's the context in which primary topic developed. --В²C 17:05, 6 August 2019 (UTC)
              • See #In the real world. Andrewa (talk) 19:44, 6 August 2019 (UTC)

In the real worldEdit

Replying to this.

Meanwhile in the real world, making main articles harder to find is not logical in the slightest. Agree. But that again begs the question. I'm arguing that deprecating P T makes articles easier to find.

We help many orders of magnitude more people when primary topics are at the base name - when the primary topic is at the base name most people get the article they want immediately, when disambiguation is primary nobody does. Disagree. That's a common assumption, and I used to make it too.

Do you really think that if the current situation was as bad as you claim that it would have lasted 18+ years? I'm not claiming that it was always this bad. Just that it is now. Consensus can change.

Do you really not recognise that if your proposal was really so much better that there wouldn't be a need to grandfather anything in? I don't think that's true. In the real world, we have existing articles and existing article names. There are good reasons not to change them. But of course these don't apply to new article names (whether they be names for new articles or new names for old ones). Yes, it's a bit radical, and "grandfathering" new articles as soon as they are created is more radical still. But logical IMO. Andrewa (talk) 20:25, 6 August 2019 (UTC)

The main factorEdit

Replying to this.

I wonder if the main factor that causes the disparate opinions about primary topic... Very important question.

,,,is what one imagines to be the target user... Exactly.

...relative to a subtle little feature that I called the search box drop down list. No. All ways of searching matter. And we all naturally want our pet ways optimised.

If your target users are those searching using that drop-down list, then totally unambiguous titles make sense. Agree of course.

If your target users are entering terms without having any feedback hints, perhaps because JavaScript is not enabled, and hitting Search (or especially GO), then ambiguous titles for primary topics make sense. Only if they agree with us on the Primary Topic. But as you say often, most will.

People imagining the dropdown-using type of target user would explain why support for primary topic seems to have lessened since the introduction of this feature. Interesting thought.

But I think we need to arrange our articles as if that feature does not exist. That's the context in which primary topic developed. Disagree. That's not the real world. But important point regarding the context of P T. And the world, and Wikipedia in particular, has now changed in many ways, not just that one. Andrewa (talk) 20:14, 6 August 2019 (UTC)

What are the other ways? --В²C 23:28, 6 August 2019 (UTC)
I guess you mean, what other ways has Wikipedia changed, and how have they affected P T? That's a complex question we don't need to answer... a tangent IMO. Far better to just acknowledge that time marches on, that's why wp:CCC, and assess where we are now... in the real world, as you or Thryduulf (can't remember which offhand) put it recently. Is that a problem? Andrewa (talk) 17:37, 7 August 2019 (UTC)
Why would they have to agree with us on PT? Ambiguous titles for primary topics make objective sense because by definition most people would want to be on the article to which they are taken. The only exception is where we choose a PT based on the cockamamie historical significance criteria and so, for example, the majority of those searching with "apple" are not taken to the (Apple Inc. article they seek.But those abominations are pretty rare and think it's reasonable to ignore them for this discussion. The vast majority of PTs are PT by the usage criteria, and so for each the majority searching with their respective titles are taken directly to the PT article they seek.
I don't understand why you objected to Thryduulf's hypothetical example. Yes, the number are made up. It's a hypothetical. But it never-the-less demonstrates how click numbers are distributed when there is a PT based on how things are laid out. You're free to tweak the numbers, but I don't think you can come up with one that demonstrates your proposal produces fewer clicks (assuming there is a genuine PT per usage method). --В²C 23:38, 6 August 2019 (UTC)
Made-up numbers don't prove or even demonstrate anything.
Why would they have to agree with us on PT? Important question. Because if they think another topic is P T, then they'll assume that when we use the base name in an article title, that other topic is the subject of the article, and won't go to it looking for the information they want. So using the base name as the article title hides that article's content for these users.
Suppose that they think that The Americans means the earlier series (ie in their opinion that's the Primary Topic), but they want the later one, and they use the search box. The dropdown list currently has no entry that they think means the series they want. The information they want is there, but we have hidden it away under titles that they assume mean something else... most notably, under the ambiguous title The Americans, which they think means the earlier series, but we think means the later one.
This is such a good specific example that I suggest we discuss at #It is already a problem above, a subsection of the already lengthy discussion specifically of The Americans. Andrewa (talk) 19:30, 7 August 2019 (UTC)
We don't base article titles solely on page stats. We all know you want us to. Andrewa (talk) 19:30, 7 August 2019 (UTC)

Replies to Thryduulf 0052 6 August 2019Edit

Thanks for this edit (although I find the edit summary a little disparaging). I think you've identified the issues well enough that I'd like to reply in detail.

Why on earth would we not want the primary topic at the base name!? That's not quite the question I asked. It is of course the main question here.

Primary topic means, by definition, that it is the topic most people are looking for, No. That view is often expressed, and has never been adopted. So it's not quite that simple. That's part of the problem.

so having the article at the title people look for it at benefits them. Disagree. That also is an important question.

It also benefits people creating links to that article, as most of them will go to the correct place whether they know the name is ambiguous or not. Disagree. The pipe trick makes linking to the unambiguous name easy. The software detects mislinkings to DABs, but can't detect them to wrong articles.

It therefore also benefits those people who fix links to disambiguation pages (as there will be fewer of them). Disagree. There will be fewer if the ambiguous term points to a DAB, and they are easier to find and fix, as the software can detect these possible mislinkings, making it possible to notify both those who create them and those who fix them.

Your comments about number of clicks are just plain wrong: There follow some figures that seem to have been pulled out of the air.

Don't forget that on a large dab page it can take some time and mental effort to find the entry you are looking for on the dab page (add another click for everyone going via a dab page organised into sections and your totals would be 900 / 500 / 2600. Now imagine the difference when dealing with the page views popular topics get Yes, and it's time well spent. Time spent loading a large wrong article is on the other hand completely wasted.

I don't know what sort of "epiphany" you have had,... Then just follow the link I gave before to User:Andrewa/my second epiphany.

but your claim that "The often-assumed benefits [of primary topics at the base name] are mythical." is directly contradictory to my nearly 15 years of experience of benefiting from primary topics being at the base name. Good point. But this experience, some of which is common to many others including most admins and page movers, was gained in the context of many assumptions, and interpreted in that context.

So, I'm asking for some lateral thinking. I know it's a challenge. It has been for me too. Thanks for being part of that process. Andrewa (talk) 21:04, 6 August 2019 (UTC)

  • I'm sorry but I'm not interested in taking this further. You seem to have developed a unique understanding of what you think primary topic means and what users want from Wikipedia that is so alien to my experience of using Wikipedia myself and of helping other people (of many different ages, skill levels, technical proficiency and subject interest) to use Wikipedia that you might as well be arguing that black is white. I've already spent more time on this than it deserves. Thryduulf (talk) 23:28, 6 August 2019 (UTC)
    Andrewa, what do you think primary topic means? --В²C 23:46, 6 August 2019 (UTC)
    I don't think that is at all clear, and I think that's an important part of the problem. It means significantly different things to different people. See #What P T means. Andrewa (talk) 18:23, 7 August 2019 (UTC)
    Sorry to lose you. I found your contributions helpful although obviously we haven't yet come to any agreement. I'm sure you are expressing what many others feel, and I'm grateful to you for taking the trouble to do that. Andrewa (talk) 18:23, 7 August 2019 (UTC)

I really wish people would read and abide by the talk page guidelines. In particular, random uses of bullet points for indenting just make it harder to reply logically, and harder to follow the strings. Or is that the whole idea? Andrewa (talk) 18:23, 7 August 2019 (UTC)

What P T meansEdit

Replying to what DO you think PT means?.

Good question. I don't think it's clear what it means. It's a technical term in linguistics (Google it) and there it means roughly what we mean by subject in English. The term is applied to other languages whose sentence grammar differs significantly from English sentence grammar, and from Latin sentence grammar etc.. When these differences are so great that they find use of the classical term subject unhelpful, some linguists use primary topic instead. Other grammarians avoid the term even in that context. IMO that's nothing to do with our usage! So it's a Wikipedian neologism.

And it has never had a clear definition. Larry Sanger seems to have invented the term as we use it, possibly without knowing of its other use, and to have subtly assumed that there was a generally agreed meaning for every ambiguous noun phrase. Bad, bad mistake in the long term. But it didn't matter much while most of the new article titles we were creating did have a generally agreed meaning, and it became a generally accepted term and principle, and lots of us (we two included) have put a lot of effort into making it work.

And as I've said many times lately, I don't think there's any reason to abandon it completely and throw all that away. Just deprecate it for new article names, whether names of new articles or new names for old articles. IMO that removes nearly all of the problems of P T, and doesn't cause much trouble at all.

As yet we have no consensus on that. Producing it is what I'm working on here. There is much still to do.

And we might even end up adopting your definition of P T as purely based on page stats. There is nothing in my proposal that contradicts that. But that's a different issue. Andrewa (talk) 18:23, 7 August 2019 (UTC)

Presumably when I wrote the above I had a diff or similar showing where the concept of Primary Topic was introduced. But I can't now find it. It was not in the original version of WP:AT, which Sanger wrote, in fact he says there Please, do not write or put an article on a page with an ambiguously-named title as though that title had no other meanings! [18] Watch this space. Andrewa (talk) 21:36, 29 August 2019 (UTC)

This version of wp:DAB mentions primary meaning which is the same idea of primary topic. It's naive... it is unlikely that there will ever be an encyclopedia article on fireplace pokers... see fireplace poker#Types. Andrewa (talk) 22:51, 29 August 2019 (UTC)

And here is where the term primary was introduced to wp:DAB. Andrewa (talk) 00:28, 30 August 2019 (UTC)

History of wp DABEdit Draft of wp:standards

More to follow! Andrewa (talk) 23:48, 30 August 2019 (UTC)


We've explored lots of tangents above, some of them very important IMO. And those side discussions can continue.

But the answer to my original question seems to be, there aren't and can't be any. The current proposal, with a few misunderstandings cleared up, would not cause any new problems whatsoever. All articles would either be easier to find, or unaffected.

Which is of course my hope and, currently, my belief also. Andrewa (talk) 05:55, 10 August 2019 (UTC)

Return to the user page of "Andrewa/P T test cases".