User talk:Andrewa/Why primary topic is to be avoided

Why this page
This leads on from User:Andrewa/Let us abolish the whole concept of primary topic. Andrewa (talk) 00:56, 12 April 2018 (UTC)

Wave (physics)
The redirect from wave (physics) was created in 2006 '' 01:31, 16 March 2006‎ Ardric47 (talk | contribs | block)‎. . (18 bytes) (+18)‎. . (from Missing encyclopedic articles) '', it would be interesting to know why as the article to which it has always redirected was (since 2003 at least) always about waves in physics.

Query posted at the creator's talk page.

Such redirects are not just harmless, they are very helpful. But AFAIK they would normally only exist as a result of a previous move. That's why this one is so interesting. Andrewa (talk) 04:29, 12 April 2018 (UTC)

Another example, see Talk:Endymion (mythology)
A fascinating RM (in which I am now involved) at Talk:Endymion (mythology).

I've opposed the move. I was tempted to let it go hoping it would take place but that would be rather pointy. So instead I've posted extensive comments on all the invalid move arguments.

So now we wait and see.

And more
An even more complex example at Talk:Charon (mythology).

Similarly Talk:Kratos (mythology).

Talk:Kotys was simpler. Andrewa (talk) 09:36, 16 April 2018 (UTC)

See also Talk:Garrison (Metro-North station). TWODABS with little claim to P T either way. Andrewa (talk) 05:46, 20 April 2018 (UTC)

Talk:Fordham station (Metro-North) and

Talk:Moyna, Purba Medinipur - both interesting P T discussions. Andrewa (talk) 20:19, 20 April 2018 (UTC)

Talk:Hypostasis (philosophy and religion)

Talk:Ross and no doubt others... Andrewa (talk) 20:22, 20 April 2018 (UTC)

Talk:Transatlanticism (disambiguation) Interesting arguments... Andrewa (talk) 07:51, 14 May 2018 (UTC)

Talk:Single Parents (TV series) seems likely to fail, although wp:SMALLDIFFS might support it. Andrewa (talk) 23:41, 18 May 2018 (UTC)

Talk:Forza (series) There aren't any obvious primary topics for the word "Forza" besides the series so it's assumed that the series is the P T. This is the second most common fallacy regarding P T, that of assuming that there must be a P T. Andrewa (talk) 15:27, 19 May 2018 (UTC)

https://en.wikipedia.org/w/index.php?title=Talk:Peter_principle&diff=841599929&oldid=840936128 Interesting !vote The concept is definitely a primary topic over the book... Is this what is meant by primary topic? Andrewa (talk) 20:25, 19 May 2018 (UTC)

Talk:Current Affairs (magazine) proposal to move to arguably ambiguous title per WP:DIFFCAPS, looks likely consensus against. Andrewa (talk) 20:29, 19 May 2018 (UTC)

Talk:Shinola discussion re interpretation of page views. Proposal was no P T but change of P T is a counter proposal. Andrewa (talk) 16:45, 20 May 2018 (UTC)

Talk:Plymouth latest RM '' A year and change after the last RM, I think we need to revisit this. The city in Devon, England is simply not the WP:PRIMARYTOPIC of the name "Plymouth"...'' Andrewa (talk) 23:55, 27 May 2018 (UTC)

Talk:Ectoplasm (paranormal) paranormal or cell biology or neither? Andrewa (talk) 23:55, 27 May 2018 (UTC)

Talk:The Force (Star Wars) looks likely to be confirmed as P T

Talk:Plymouth A year and change after the last RM, I think we need to revisit this... and a very messy RM discussion in which I got a mention Andrewa (talk) 17:14, 29 May 2018 (UTC)

Possible test case
Talk:Magic (paranormal), current RM is the second this year, at least two previous moves, DAB (see Magic (disambiguation)) has 154 (that's one hundred and fifty-four) entries by my IEHEYEBALL count. Andrewa (talk) 16:43, 28 May 2018 (UTC)

The Man in the Gray Flannel Suit
Talk:The Man in the Gray Flannel Suit (novel) is interesting... the novel is probably more significant than the film. But better to disambiguate both. Might be a better example than wave. Andrewa (talk) 04:41, 9 June 2018 (UTC)

Unsure
Not sure what these examples prove

https://en.wikipedia.org/w/index.php?title=Talk:Sex_Bomb_(song)&oldid=845837150 (two RMs, one open)

https://en.wikipedia.org/w/index.php?title=Talk:Tranny_(slang)&oldid=845971217#Requested_move_11_June_2018 (still open)

but both interesting. Andrewa (talk) 23:55, 18 June 2018 (UTC)

Breaking external links
https://en.wikipedia.org/w/index.php?title=Talk:Kim%27s_Convenience_(TV_series)&diff=845313915&oldid=832318660 Interesting? Andrewa (talk) 00:11, 19 June 2018 (UTC)
 * https://en.wikipedia.org/w/index.php?title=Talk:Kim%27s_Convenience_(TV_series)&diff=846639770&oldid=846571755 Closed as DAB at base name. Andrewa (talk) 03:29, 20 June 2018 (UTC)

https://en.wikipedia.org/w/index.php?title=Talk:Dominance_and_submission&oldid=847907861#Requested_move_20_June_2018 Possible P T, DAB existed but there was no otheruses hatnote so I added it. Andrewa (talk) 16:46, 28 June 2018 (UTC)

https://en.wikipedia.org/w/index.php?title=Talk:Sarah_Champion_(politician)&oldid=847347805#Requested_move_24_June_2018 and lots more

Boater
See Boater (disambiguation) and Talk:Boater... is this a classic example in which current rules give a P T but it would be in the readers' interest if the DAB had been at the base name when first created? Andrewa (talk) 17:59, 4 July 2018 (UTC)

Now some support for having the DAB (which now has been expanded by another two entries making four in all) being at the base name, but looks unlikely to happen. Andrewa (talk) 21:30, 4 July 2018 (UTC)

Other essays

 * Wikipedia:Primary topics in WP:TWODABS situations

Great hypothetical examples, but the assumption is that mouse clicks are all that matter.

What others I wonder? Andrewa (talk) 05:48, 20 April 2018 (UTC)

Phones
My phone does not give a drop down list of suggested destinations... 

''Regarding search on a phone, I'm no expert, but I tried using my old phone which has several browsers installed. One showed me a desktop version, which when zoomed in so I could see the tiny letters, worked just like the desktop with a drop-down list. On another browser the search button didn't work at all, and on a third typing Plymouth in the search box gave no drop-down, but on pressing enter I was taken to a page listing the possible matches. So on this tiny sample, it seems possible that your phone is using a browser that's not compatible with WP's enhanced search function.''

Very interesting; It seems that mobile users do get a search results list after all. Andrewa (talk) 11:44, 1 June 2018 (UTC)

National variations
To make it worse, there are substantial national effects - what is "more significant" to a reader is biased by their nationality

''Well stated. It's certainly true that in this case readers from different countries will have very different experiences with which is the most historically significant topic. For American readers, Plymouth, Massachusetts/Plymouth Colony is absolutely the primary topic. For people in Montserrat, it's Plymouth, Montserrat. For Canadians, it may be the car, etc. Plus, the second part of the guideline was never meant to be read in isolation from the first part. This isn't a case where some new pop culture topic beats out a more significant topic in page views for awhile; all those topics have substantial long-term significance. When there's this level of legitimate variance, disambiguation should be the default option.'' (the first part of this edit also worth a read)

Interesting views, specifically related to Plymouth but perhaps with wider application. Andrewa (talk) 06:33, 31 May 2018 (UTC)

History
A dab page would be the proper landing, can you imagine a written encyclopedia which decides where you should go first before letting you look at the index?

Can any of us imagine a written encyclopedia that maintained gaps in its index because of a decision an influential but now-retired editor made in the days when the encyclopedia was copied by hand rather than printed? But ISTM that's the position we are taking with regard to primary topic! See User:Andrewa/Why primary topic is to be avoided. Andrewa (talk) 14:52, 2 June 2018 (UTC)

As that section (User:Andrewa/Why primary topic is to be avoided) has developed, this has begun to seem a perennial proposal candidate! (But no, it is not listed there, not yet at least.) It still seems to me to be due for a re-evaluation.

I have posted a talk page comment and sent an email to the editor who introduced the term primary meaning, but they have not edited for two years now. Andrewa (talk) 15:59, 2 June 2018 (UTC)

Some excellent points
has been making some excellent points with regard to this essay, but on my main talk page.

Here are the latest, and some replies.

With regards to the last point, I don't think scrapping PT is viable. I don't think it is currently viable. I suspect it will be the eventual result, and that the sooner we do it the better in view of this. But it may be long after I have ceased to edit.

Moving the goalposts substantially to would make the current battlefield cases non-issues, sure. Yes, and would have other less obvious benefits which are more important and a bit surprising.

''But not outright scrapping. That will affect many vital articles, like those on every major academic subject and every country. Due to the extreme importance of those subjects, they are also likely to have the most incoming links and those links would be trashed.'' No, these links would not be trashed. Those that are accurate would require one more mouse click.

''IMO its natural that WP developed PT. In general, the most important subjects of the name were the first ones to have had the article created. Those became PT by default, not by conscious editorial decision-making. The PT rules are a codification of the that, and a way to address those anomalies where the wrong subject is at the base name.'' Agree. Well put.

Trying to eradicate PT might solve the current issues, but would make matters worse for much more important articles. Disagree. That's the whole point of User:Andrewa/negative benefit.

It is a common and natural assumption. I made it myself until recently. But it doesn't stand up to scrutiny. That surprised me and I think will surprise almost everyone.

But just to summarise, I'm not currently advocating that we scrap P T. I am: Thanks for your excellent contributions to both of these aims. Andrewa (talk) 21:38, 10 June 2018 (UTC)
 * Strongly advocating that we raise the bar considerably.
 * Trying to raise awareness that the long and generally held assumption that it's to the readers' benefit to have important articles at ambiguous names is highly questionable.
 * With regards to the bit you "Disagree": You are only considering those people who land on WP pages by searching. Searching is not the only access method to WP. You need to factor in those who follow links from other places. Those readers will get a negative impact, and if a link from an school-level education website targeted at Mathematics gives the content currently at Mathematics (disambiguation), not only does that add an extra click to the students using it, it makes us look silly. Taking the long view (that those links will repair themselves with time) is all well and good, but isn't realistic - books will not be re-printed, historical websites will not be updated, while well-maintained websites will be slow to respond and only change if their users give them feedback. Please acknowledge link-followers into your essay - they are negatively affected by changing from "there is a primary topic" to "there isn't a primary topic". It currently gives the impression there are zero drawbacks to all users - NOT true.
 * I agree with your bulleted conclusion points.--Nilfanion (talk) 22:34, 10 June 2018 (UTC)
 * Yes, I agree that the matter of these external links etc should be addressed.
 * But I do not agree that this makes Wikipedia look silly... or at least, we can't please everyone, and some will see whatever we do as silly. But we look far sillier at present, in my judgement. It's probably a very personal call. Andrewa (talk) 06:12, 11 June 2018 (UTC)
 * Silly is hyperbole of course. But what I mean is that the current arrangement for mathematics doesn't appear odd, even if it may hinder some. That's because no-one looking for one of the other mathematics would be remotely surprised to get land on the concept involving numbers. But switch the situation and it would seem strange, as the people hindered are likely to have not heard of the minor terms, and wonder why they are having to jump through hoops to get to the "obvious".--Nilfanion (talk) 18:00, 11 June 2018 (UTC)
 * I'm not sure it's even hyperbole. It's a valid point. But is it, on balance, a pro or a con? Someone looking for London, England could be very surprised to find that the term is ambiguous, and if about to book an air ticket, very relieved to discover in time that the bargain fare they had found online didn't actually cross the Atlantic (it has been reported in the news that others have been not so fortunate... and travelers to Sydney and possibly Paris have occasionally had the same problem, Sydney definitely in real life, Paris in a Carl Barks story at least). Someone looking for Mathematics might well consider Wikipedia stupid for even having some of the other articles, but IMO that is because they are themselves stupid (and Mathematics is arguably the love of my life). We don't need to follow their poor logic. Most people would be interested to discover that it's also a magazine title, even if they didn't have much interest in the songs. Andrewa (talk) 15:25, 13 June 2018 (UTC)


 * Regarding the point about replacing an article with a DAB and forcing readers to make an extra click making us look silly. What about in situations where we delete a long standing article, when the reader sees "this page has been deleted" that makes us look silly, but I don't see how this has the same effect with primary topics apart from the fact that mathematics is a major topic and many deleted articles aren't.  Crouch, Swale  ( talk ) 07:35, 12 August 2018 (UTC)

Mor Maman
There's a concrete example in the recent RM at Talk:Mor Maman.

The article on the the footballer was created in 2006. Then in 2014, an article was created for the (now former) beauty queen. 

Now in 2018, four years later, we are likely to move the article on the footballer to an unambiguous name. This will, as you say, inconvenience (slightly) all those who follow links to this article created in those four years... and I will point out, this also includes wikilinks within English Wikipedia unless someone fixes them, and there is at present no process that requires this although it does sometimes happen and we could easily institute one (creating a little more work for editors too).

So, in this case (and assuming the move goes ahead) there would have been a reader benefit in this move occurring four years ago, when it was first known that the title of the article on the footballer was ambiguous. (And there would also have been several benefits to editors.)

, interested in your comments. Andrewa (talk) 15:21, 11 June 2018 (UTC)
 * The potential benefits to readers, which you describe well in your essay, would have applied from any point - even before the beauty queen's article was created. When a title like that is moved away from the base name, the internal links should be easy enough to identify. There should be some requirement on the person who closes an RM to list the move somewhere, so the gnomish editors (or their bots) can fix the links.
 * I think the benefits of moves that remove PT status need to be balanced against the link-preserving benefits of the status quo. I don't have any real idea of how to balance those two factors, as I don't think we have the information needed (such as the % of users who get to these articles via various routes). I think that needs some editor thought on suitable proxy measures to check and tools to use, and those tools will be different to the current PT ones. For instance while article age is irrelevant to PT, it is crucial to stability - the longer an article is at a title, the more valuable it is to keep it there.
 * My guess is that as the subject becomes more important, the benefits of the status quo increase. The readers of Mor Maman don't really gain much from stability. But the readers of Israel do.--Nilfanion (talk) 17:52, 11 June 2018 (UTC)
 * Agree with the statements of fact and even most of the opinions but dispute the methodology.
 * Two problems with the methodology.
 * One is, we don't always need empirical data. It's a good thing to use it when it's available and even to rely on it when it can be interpreted unambiguously. But analysis is also valuable and most often required.
 * The other is, we need to compare apples to apples. It doesn't in the first cut matter whether benefits to readers of less important topics are greater or less than those concerned with more important topics. What I want to do first is consider benefits to all articles. If there are benefits all around (and I think that there are, which as I've said surprised me and I think would be surprising to most) then which topics are more or less important does not matter.
 * And that's the basic mistake that has been made with respect to primary topic. We've focused on the more important articles, assuming (correctly) that the overall reader experience will be best if they get favourable treatment... or in other words, giving a positive loading to the data concerning readers of these articles. But this mindset was, in hindsight, not good for sound analysis. If we'd focused instead on the minority readers looking for relatively obscure topics, we'd have made a better decision for the majority readers as well.
 * And I think I may have just invented a seventh coloured thinking hat. This is essentially the same principle as my favourite maxim of software debugging... fix the unimportant problem you do understand, and the important problem you don't understand will often just go away too.
 * Agree particularly that the benefits of moves that remove PT status need to be balanced against the link-preserving benefits of the status quo. Important point. Andrewa (talk) 21:01, 11 June 2018 (UTC)
 * Agree we don't need hard data all the time of course, the problem is at this point its not clear what data/facts is available, never mind if its helpful or not.
 * My comments regarding importance are based on my belief that there isn't a directly proportional relationship here (ie a page with ten times as many page views attracts ten times as many readers via external links). Rather, I believe that traffic pattern is different for the more significant articles, and they get a higher proportion of their traffic from external links. If I'm correct, those pages need a higher level of "status quo protection", and therefore articles on major subjects need different treatment to those on minor subjects.--Nilfanion (talk) 21:31, 11 June 2018 (UTC)
 * Agree it's likely (safe to assume even) that traffic pattern is different for the more significant articles, but not convinced that they need different treatment to those on minor subjects. They may, but it's not to be taken lightly, and one problem is, where do we draw the line? The balance that needs to be struck is between the (IMO small) impact of looking silly to some versus the far greater (IMO) impact of easier navigation for readers and avoiding a massive time sink for editors and administrators. Andrewa (talk) 23:44, 11 June 2018 (UTC)
 * Its not the "looking silly" that's the problem, but the impact to those readers who have followed a link and are expecting to get directly to an article (ie the additional mouse click). The cost of that, per error, is lower than the cost per error by landing on a PT incorrectly; but the total cost is likely to be substantial on those articles with a lot of incoming links. Agree "where to draw the line" is a problem, and that's a major reason for wanting to find some suitable numerical measure. Its hard to do a meaningful cost/benefit analysis without some guiding figures, and pure discussion will lead to trouble on articles in the middle... like Plymouth! Avoiding the problem by insisting on a single approach for everything isn't workable, so some line needs to be drawn somewhere.--Nilfanion (talk) 07:17, 12 June 2018 (UTC)
 * My ideal with-hindsight solution would be that adopted by Wikidata: Breaking the link between the URL and the article's title. The URL can be a permanent fixed identifier, while the title can be a suitable name. That would both eliminate all linkrot and completely avoid all clashes between two subjects with the same name.--Nilfanion (talk) 21:38, 11 June 2018 (UTC)
 * Interesting idea re breaking the link between displayed title and URL. I'm not familiar with Wikidata and my experiences haven't been good to date! But it sounds like a good approach and perhaps it's not too late to adopt something similar... it could be implemented as an optional parameter that defaulted to having them the same as now. But it doesn't eliminate the problem of people wanting to promote a particular agenda by using Wikipedia's undoubted and perhaps unfortunate perceived and practical authority (see Macedonia), and might even aggravate it. I want to think about that one! Andrewa (talk) 23:44, 11 June 2018 (UTC)
 * One thing which could be adopted is to make the search behave similarly. Instead of just giving a list of titles, WD's search gives title + one sentence description. It is also more capable as it searches the entire title instead of the start. Macedonia is a good example: On WP, I can't even see the country on the drop-down. On WD the top options include: "Republic of Macedonia (Macedonia) - a country in southeastern Europe", "Macedonia - Roman province", "Macedonia - a geographical region of Greece".--Nilfanion (talk) 07:19, 12 June 2018 (UTC)

.you might like to check my logic here, as it's based on your arguments above (as I think I understand them).

Some good points above... thinking hard about them. Andrewa (talk) 11:37, 12 June 2018 (UTC)

Progress
, I've just updated the essay to incorporate one of your excellent points above... comments? Andrewa (talk) 15:07, 13 June 2018 (UTC)
 * That looks good to me. However one minor issue, which I don't want to dwell on, is that finding an unambiguous name can be harder than you think - so a move from an ambiguous name might end up going to another ambiguous name, despite the good intentions. Geography provides good examples for this, and the naming conventions show this.
 * As examples both WP:UKPLACE and WP:USPLACE provide three tiers of disambiguation. County, then district, then something else for the UK (eg Woolston, south Shropshire); and state, then county, then something else for the US.
 * In those cases the natural inclination is to stop as soon as possible, and when forced to move a topic down it might not go down far enough and still be ambiguous. Luton, Devon seems fine, but there are two Lutons in the county. There are dozens of similar examples.
 * What it certainly is is a very good reason to avoid creating new primary topics.--Nilfanion (talk) 17:30, 13 June 2018 (UTC)

Difficulty identifying ambiguous names
makes a good point above with regard to the difficulty of finding an unambiguous name. The point seems to be, a name that appears unambiguous can later turn out to be ambiguous, when another article is created that uses the same base name and perhaps even the same higher level disambiguator(s). Luton, Devon is an excellent example. There is a Luton in East Devon http://www.visionofbritain.org.uk/place/24942 but it is not even clear to me whether or not that is the topic of our current stub! Andrewa (talk) 22:57, 2 July 2018 (UTC)