User talk:Andrewa/advantages and disadvantages of ambiguous article titles

Why this page
This was inspired by this contribution to an RM. It got me thinking, there are other reasons, but have I ever seen them listed succinctly? Andrewa (talk) 23:50, 21 October 2019 (UTC)

An interesting essay
Wikipedia:Primary topics in WP:TWODABS situations gives some interesting hypotheticals.

It may be worth reinterpretting them to see whether they support the deprecation of Primary Topic. Certainly worth a closer look. Andrewa (talk) 02:34, 22 October 2019 (UTC)

Another advantage of primary topics: cleaner wikitext
There are currently 120,649  wikilinks in mainspace articles. Imagine they all needed to be replaced with. Yuck. Unpiped links are easier to write and make wikitext source shorter and easier to read. Colin M (talk) 03:11, 22 October 2019 (UTC)
 * Thanks for that contribution. . But this edit subtly changed the format. Peviously, bolding was used only on the first level (numbered) points, to give the main issues prominence. The bullets are more of the way of explanation. Would you mind if I reformatted your point, just adjusting the bolding, to comply to this? Andrewa (talk) 02:25, 24 October 2019 (UTC)
 * By all means! Colin M (talk) 05:38, 24 October 2019 (UTC)
 * Agree that changing all the wikilinks to France is a very silly idea. If we were starting Wikipedia from scratch it might be different, but we're not. France should stay where it is, and the wikilinks won't need changing.
 * And it might not be any different even if starting from scratch. Even in the case of a newly ambiguous term, once the article has been created at a (then) unambiguous name, I think it should be able to keep that name, even if it later becomes ambiguous. The article should only be moved if there's consensus that it is not the primary topic. In that case there is no choice... just as now. And it would be a hassle! Just as now.
 * In the case of France it seems unlikely. Andrewa (talk) 02:42, 24 October 2019 (UTC)
 * I have a lot of thoughts about the grandfathering clause in your primary topic RfC, but for now I'll just note that this essay is currently framed as "advantages and disadvantages of ambiguous article titles", and "France" would certainly qualify as one. Colin M (talk) 05:47, 24 October 2019 (UTC)
 * Exactly. You said Imagine they all needed to be replaced with . That's irrelevant. What we should be imagining is that they'd all been created with this code in the first place. And that's a very different scenario, and even seems to be what Larry Sanger had in mind many years ago.
 * And even that doesn't really do the thought experiment justice. In the case of Paris, maybe almost everyone knows it's ambiguous. Maybe. There's a plot device in an Uncle Scrooge comic based on Gladstone Gander not knowing that Paris is ambiguous, and being sent to Paris, Texas instead of Paris, France (and note that even with Paris, France the pipe trick isn't always necessary or even desirable). That indicates to me that Carl Barks and his readers found it quite credible that many Americans would not even know that Paris was ambiguous. Otherwise it wouldn't be funny.
 * My conclusion is that Paris is an outrider... likely to be the occasional exception, and needing to be treated with commonsense as a result. Andrewa (talk) 13:12, 25 October 2019 (UTC)

This brings us back to the basic question. Currently, we have a DAB at New York but an article at Paris. Why? What's the difference?

Aren't our readers and contributors in much the same position with both cities? Most will assume that both New York City and Paris, France are what they'll get to when they go to or link to New York and Paris respectively. But some will not. We currently give some help to those who go to or link to New York expecting the state, but none to those who make a similar mistake with Paris.

And with Paris there's no fixing that, and it's not a big problem anyway. But where new and/or less clear-cut ambiguities arise, shouldn't we go with the New York solution and default to a DAB? Andrewa (talk) 17:15, 26 October 2019 (UTC)

Contesting wording in advantage 1
Advantages #1 says: Anybody who wants the article which is at the ambiguous name and doesn't know that the title is ambiguous may be saved a few keystrokes (bold added)

I think the bolded part is inaccurate. When I search for "France" or "Snail" or "Marx", expecting to be taken to a specific article (about a country/invertebrate/philosopher), it's not because I don't know that those names are ambiguous. It's because I'm confident that the topics I'm seeking are so significant and so strongly associated with those terms, that Wikipedia will be able to correctly guess what I'm looking for. Colin M (talk) 16:01, 23 October 2019 (UTC)
 * If you know that the title is ambiguous, you're not one of the people in question. That advantage is only about people who don't know that the title is ambiguous. And there will be some of these, for some titles at east. I don't see anything unclear, let alone inaccurate.
 * What you seem to be saying is that someone who is confident that the topics I'm seeking are so significant and so strongly associated with those terms, that Wikipedia will be able to correctly guess what I'm looking for is saved keystrokes too. Very good point. Have a go at wording it succinctly and feel free to add it. Once you've done that we can see whether the two points might be merged.
 * Or you could try to reword my point to cover both cases. But I'm not suggesting that, as you don't seem to understand the point I was making. Andrewa (talk) 03:00, 24 October 2019 (UTC)
 * What about just removing the bolded text? Since not knowing that the title is ambiguous is neither a necessary nor a sufficient condition for saving keystrokes. Colin M (talk) 05:50, 24 October 2019 (UTC)
 * See . Andrewa (talk) 16:48, 26 October 2019 (UTC)

The basis of the analysis
When Wikipedia has chosen a P T (which implies we take the term to be ambiguous of course), there are four possible scenarios: Part of my thinking is that we can clarify things by considering these four cases one at a time. That's the function of the text doesn't know that the title is ambiguous.
 * 1) The reader also thinks that the title is ambiguous
 * 2) ... and agrees with us on the P T
 * 3) ... but doesn't agree with us on the P T
 * 4) They may think that there's another P T
 * 5) They may not think there is a P T at all
 * 6) The reader doesn't think the title is ambiguous
 * 7) ... and agrees with us on its meaning (trivially, they agree that our meaning is the P T)
 * 8) ... but disagrees with us on the meaning.

These are all interesting. 1.2 and 2.2 are the difficult ones under our current system. The question is, do its advantages in cases 1.1 and 2.1 justify the confusion of 1.2 and 2.2? I'm obviously skeptical. But very interested in the investigation, and grateful for input from those who see the other side of this.

We can't do the analysis the other way around, which is what we're trying to do if we delete doesn't know that the title is ambiguous. This lumps together two very different scenarios. Andrewa (talk) 16:53, 26 October 2019 (UTC)