A few iterations of the inclusionism/deletionism debate back, we seem to have settled on merging articles as a sort of happy medium. Increasingly, though, it seems to me that mergism and redirectionism is proving more destructive to our content and its growth than deletion was.
I'll limit myself to fiction articles, since that's where I've seen the worst effects, though I'd love to hear from people who edit in other areas. [[List of characters in Gilmore Girls]] was the target of a wealth of merges of characters, such that no characters in the show have individual articles anymore. And, indeed, the old character articles were crappy in-universe messes of the sort we want to clean up.
The problem is that it is easy for any of the major series regular characters to have an article written about them. Gilmore Girls was a critical darling of a show, actors regularly gave interviews, all seasons are on DVD with a decent number of special features that provide out-of-universe information. The information is clearly and transparently there. The articles could have eventually been improved.
But the articles did not satisfy notability in their old forms, and so are now gone. And, worse than gone, they're redirects - which means that a newbie user is going to have a much harder time figuring out how to go about fixing them. Redlinks at least cry out to be fixed. Redirects avoid being fixed. And since the characters now exist in a list, incremental improvement is a real challenge. The format of the articles doesn't lend itself to expansion into new areas, as it seems weird to have only one entry on a list have out-of-universe information. Furthermore, the nature of a list os succinctness - expanding an entry with a lot of information is unwanted.
Deletion at least left a visible hole in our coverage that anybody could see and fix. Redirects, through a combination of unclear interface ("How do I fix/make a redirect" is just about the most common question asked by my non-wiki using friends when they try to edit) and an institutional resistance to un-merging that is almost as bad as abusive G4 speedies, redirectionism seems to me to destroy our coverage of areas more severely than deletion.
The real culprit here is WP:N, which does not do nearly enough to protect articles on topics that obviously could pass its standard of notability, but do not yet. The anti-eventualist bias of this requiring of multiple independent sources to be cited before an article can avoid deletion is appalling. We need to remember that articles grow slowly, and that a mediocre start is still a better foundation for an article than a redirect.
-Phil
On Tue, Apr 15, 2008 at 1:58 PM, Philip Sandifer snowspinner@gmail.com wrote:
A few iterations of the inclusionism/deletionism debate back, we seem to have settled on merging articles as a sort of happy medium. Increasingly, though, it seems to me that mergism and redirectionism is proving more destructive to our content and its growth than deletion was.
I'll limit myself to fiction articles, since that's where I've seen the worst effects, though I'd love to hear from people who edit in other areas. [[List of characters in Gilmore Girls]] was the target of a wealth of merges of characters, such that no characters in the show have individual articles anymore. And, indeed, the old character articles were crappy in-universe messes of the sort we want to clean up.
The problem is that it is easy for any of the major series regular characters to have an article written about them. Gilmore Girls was a critical darling of a show, actors regularly gave interviews, all seasons are on DVD with a decent number of special features that provide out-of-universe information. The information is clearly and transparently there. The articles could have eventually been improved.
But the articles did not satisfy notability in their old forms, and so are now gone. And, worse than gone, they're redirects - which means that a newbie user is going to have a much harder time figuring out how to go about fixing them. Redlinks at least cry out to be fixed. Redirects avoid being fixed. And since the characters now exist in a list, incremental improvement is a real challenge. The format of the articles doesn't lend itself to expansion into new areas, as it seems weird to have only one entry on a list have out-of-universe information. Furthermore, the nature of a list os succinctness - expanding an entry with a lot of information is unwanted.
Deletion at least left a visible hole in our coverage that anybody could see and fix. Redirects, through a combination of unclear interface ("How do I fix/make a redirect" is just about the most common question asked by my non-wiki using friends when they try to edit) and an institutional resistance to un-merging that is almost as bad as abusive G4 speedies, redirectionism seems to me to destroy our coverage of areas more severely than deletion.
The real culprit here is WP:N, which does not do nearly enough to protect articles on topics that obviously could pass its standard of notability, but do not yet. The anti-eventualist bias of this requiring of multiple independent sources to be cited before an article can avoid deletion is appalling. We need to remember that articles grow slowly, and that a mediocre start is still a better foundation for an article than a redirect.
Indeed, Phil. I've become more certain over the years that generally, the question of whether there should be an article, a redirect, or no article for a particular topic should be dependent alone on the nature of the topic, and not the nature of the article we happen to have about it at the time. If George W. Bush is notable, there can be little excuse for deleting or redirecting his article to [[Presidents of the United States]] just because, say, the first revision ever was "he eatz poop" or even something more coherent blaming him for every evil on the planet. Likewise for TV shows, and so forth.
What concerns me is that we often have a lower standard for mergers and redirections than we do for deletion. That only makes sense, but it does sometimes lead to strange situations. I still remember someone once proposed merging and redirecting our articles on the lower and upper houses of the Malaysian Parliament to [[Parliament of Malaysia]] because there was supposedly insufficient content in those two articles to merit keeping them separate. Even if we took the latter to be true, it is disturbing that there is rarely any consideration of whether the article topic is in itself notable.
A good first step nevertheless would be to make undoing redirections easier. At the moment only experienced users can handle this. I am not sure about how we can best modify the user interface to make fixing redirects more transparent, while not annoying casual readers, but we should start thinking about this problem now; it's way overdue to be fixed, come to think of it.
Johnleemk
On Tue, Apr 15, 2008 at 1:58 PM, Philip Sandifer snowspinner@gmail.com wrote:
I'll limit myself to fiction articles, since that's where I've seen the worst effects, though I'd love to hear from people who edit in other areas. [[List of characters in Gilmore Girls]] was the target of a wealth of merges of characters, such that no characters in the show have individual articles anymore. And, indeed, the old character articles were crappy in-universe messes of the sort we want to clean up.
There are a number of problematic *types* of merges, and I think you've hit on one of them. To my mind it never makes sense to redirect an instance of something to a list. Certainly not an article which clearly designates itself as a list, and more arguably not to those listy articles whose title doesn't explicitly call itself a list.
Of course, I see lists more as navigation tools than as articles in themselves, whereas the standard practice seems to mix articles and lists together. Surely it's in large part due to the fact that making lists is much easier than making an article which provides an overview of the topic. I also think the nature of wiki-collaboration leads to this type of article development. So it's probably a difficult problem to fix.
On the other hand, I just hit random page a dozen or so times and couldn't find any instances of these list-like articles. So I don't think it's for a large percentage of articles, though I do suspect a sample weighted by article traffic would show a bigger problem.
And then there are articles like [[YouTube]], which is almost the opposite problem (not enough merging - there are two "sections" which consist entirely of ''Main article: [[whatever]]''). Although, viewed differently, maybe it's the same problem - lack of synthesis.
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On Wed, Apr 16, 2008 at 8:54 AM, Anthony wrote:
On Tue, Apr 15, 2008 at 1:58 PM, Philip Sandifer wrote:
I'll limit myself to fiction articles, since that's where I've seen the worst effects, though I'd love to hear from people who edit in other areas. [[List of characters in Gilmore Girls]] was the target of a wealth of merges of characters, such that no characters in the show have individual articles anymore. And, indeed, the old character articles were crappy in-universe messes of the sort we want to clean up.
There are a number of problematic *types* of merges, and I think you've hit on one of them. To my mind it never makes sense to redirect an instance of something to a list. Certainly not an article which clearly designates itself as a list, and more arguably not to those listy articles whose title doesn't explicitly call itself a list.
Of course, I see lists more as navigation tools than as articles in themselves, whereas the standard practice seems to mix articles and lists together. Surely it's in large part due to the fact that making lists is much easier than making an article which provides an overview of the topic. I also think the nature of wiki-collaboration leads to this type of article development. So it's probably a difficult problem to fix.
On the other hand, I just hit random page a dozen or so times and couldn't find any instances of these list-like articles. So I don't think it's for a large percentage of articles, though I do suspect a sample weighted by article traffic would show a bigger problem.
And then there are articles like [[YouTube]], which is almost the opposite problem (not enough merging - there are two "sections" which consist entirely of ''Main article: [[whatever]]''). Although, viewed differently, maybe it's the same problem - lack of synthesis.
WikiEN-l mailing list WikiEN-l@lists.wikimedia.org To unsubscribe from this mailing list, visit: https://lists.wikimedia.org/mailman/listinfo/wikien-l
As a deletionist, I must admit I do agree that we should look at the topic of an article rather than its content. In some cases (A7, if I can remember CSD correctly--the one that deals with notability), the content is the only data to go on, especially for esoteric articles. Ghits are often used as a gauge when information is not available/known (e.g. if someone creates a nn bio of themselves or their friends or some other equally obscure topic). Often, if an article reads like it's just some random person, it gets deleted. But I digress. We're talking about merges. A merge often happens when there is /insufficient/ content. If the content hints at nonnotability, it is usually deleted. We should better encourage unmerging (is that a word...?), and create a /simple, easy to use/ system/gui for doing so. However, merging itself is not evil. If an "article" (a stub) is two sentences long, it makes more sense to group it with related information. That way, we (as a community) don't need to maintain an increased number of articles (yes, they still exist, they still take up space, but we don't need to protect them from vandalism etc., we don't need to update them as e.g. external links change, and a lot more), and the reader gets to read more than a few sentences. We presume that by entering a topic, they wanted information about it (or they wanted to edit it, but at least /some/ will want to read). Ergo, if there's only a few sentences on it available, they will (probably) want related/more information/external links, which a list provides.
I'm sorry to break off in the "middle" (end? I don't know) of a [[stream of consciousness]], but, well, let's just say something happened in the real world. Not a major crisis (or indeed something I seriously need to worry), but more than a distraction (or more than the average distraction). Not anything to worry about, so please don't panic
- - - -- Sincerely, [[User:Thinboy00]]
On Thu, Apr 17, 2008 at 4:59 PM, [[User:Thinboy00]] thinboy00+wikipedialist@gmail.com wrote:
If an "article" (a stub) is two sentences long, it makes more sense to group it with related information. That way, we (as a community) don't need to maintain an increased number of articles (yes, they still exist, they still take up space, but we don't need to protect them from vandalism etc., we don't need to update them as e.g. external links change, and a lot more), and the reader gets to read more than a few sentences. We presume that by entering a topic, they wanted information about it (or they wanted to edit it, but at least /some/ will want to read). Ergo, if there's only a few sentences on it available, they will (probably) want related/more information/external links, which a list provides.
This is, generally, the essence of the argument for merging, and I find it completely without merit. I have on more than one occasion found myself on one of these megamerged pages, and each time I would have preferred a two sentence stub which linked to the list page. Either way all I'm really getting is two sentences of useful information, but at least with the stub my browser loads quicker and I don't have to search for those two sentences. You say "the reader gets to read more than a few sentences", but if the rest of those sentences are about something other than what I'm searching for, that's not a positive thing.
I don't buy that it's easier to maintain these megamerged pages. If anything I would think it would be easier to hide vandalism among the more frequently edited megapage - plus external links still need to be maintained, what links here stops working, categories don't work, editing is harder, etc. Would it be easier to maintain [[Arthuriana]], [[Crustaceana]], and [[Infection and Immunity]] if we merged them together into [[List of minor scientific journals]]?
On Apr 17, 2008, at 8:37 PM, Anthony wrote:
This is, generally, the essence of the argument for merging, and I find it completely without merit.
It's not without merit as such... it's just that its merit seems based primarily on the logic of 1.0, which, while an interesting project when it started, grows more questionable by the day, both in terms of how helpful it will be and how likely it is to happen.
I don't particularly mind merging perma-stubs, but an article that could develop should be allowed to.
-Phil
[[User:Thinboy00]] wrote:
A merge often happens when there is /insufficient/ content. If the content hints at nonnotability, it is usually deleted. We should better encourage unmerging (is that a word...?), and create a /simple, easy to use/ system/gui for doing so. However, merging itself is not evil. If an "article" (a stub) is two sentences long, it makes more sense to group it with related information. That way, we (as a community) don't need to maintain an increased number of articles (yes, they still exist, they still take up space, but we don't need to protect them from vandalism etc., we don't need to update them as e.g. external links change, and a lot more), and the reader gets to read more than a few sentences. We presume that by entering a topic, they wanted information about it (or they wanted to edit it, but at least /some/ will want to read). Ergo, if there's only a few sentences on it available, they will (probably) want related/more information/external links, which a list provides.
I'm not sure I really see the improvement though. If anything, these big list articles are much harder to organize, edit, and keep decent than smaller, focused articles are. As far as I can tell the main impetus for them is really a feeling that not-particularly-famous characters in some not-particularly-famous universe don't really "deserve" their own articles, and that merging them all into one somehow reduces clutter.
I'd guess that sort of embarrassment about articles on topics deemed not particularly noteworthy is why it comes up almost exclusively with fictional characters---you don't see anyone pushing to merge 500 minor Roman consuls into a big article like [[List of Roman Republic consuls about whom history has recorded no more than about 4 sentences worth of information]], or lots of very small towns in Minnesota into [[List of Minnesota towns with fewer than 100 people]]. They just each get their own, short article. This is more convenient for a number of reasons; for example, the Roman consuls who lived in the 3rd century BC can be conveniently found via [[Category:3rd century BC people]], which would be hard to do if they were all in a big list.
-Mark
On Thu, Apr 17, 2008 at 11:47 PM, Delirium delirium@hackish.org wrote:
[[User:Thinboy00]] wrote:
A merge often happens when there is /insufficient/ content. If the content hints at nonnotability, it is usually deleted. We should better encourage unmerging (is that a word...?), and create a /simple, easy to use/ system/gui for doing so. However, merging itself is not evil. If an "article" (a stub) is two sentences long, it makes more sense to group it with related information. That way, we (as a community) don't need to maintain an increased number of articles (yes, they still exist, they still take up space, but we don't need to protect them from vandalism etc., we don't need to update them as e.g. external links change, and a lot more), and the reader gets to read more than a few sentences. We presume that by entering a topic, they wanted information about it (or they wanted to edit it, but at least /some/ will want to read). Ergo, if there's only a few sentences on it available, they will (probably) want related/more information/external links, which a list provides.
I'm not sure I really see the improvement though. If anything, these big list articles are much harder to organize, edit, and keep decent than smaller, focused articles are. As far as I can tell the main impetus for them is really a feeling that not-particularly-famous characters in some not-particularly-famous universe don't really "deserve" their own articles, and that merging them all into one somehow reduces clutter.
I'd guess that sort of embarrassment about articles on topics deemed not particularly noteworthy is why it comes up almost exclusively with fictional characters---you don't see anyone pushing to merge 500 minor Roman consuls into a big article like [[List of Roman Republic consuls about whom history has recorded no more than about 4 sentences worth of information]], or lots of very small towns in Minnesota into [[List of Minnesota towns with fewer than 100 people]]. They just each get their own, short article. This is more convenient for a number of reasons; for example, the Roman consuls who lived in the 3rd century BC can be conveniently found via [[Category:3rd century BC people]], which would be hard to do if they were all in a big list.
Mmm, precisely. Indeed, in the case of the Malaysian Parliament I mentioned earlier, there was actually enough content in the articles for both houses of Parliament for them not to be labeled stubs; the main reason cited for proposing a merge, as I recall it, was that they were just not all that notable, and could and should be merged into the main article on Parliament itself. Perhaps in theory merges are only for content that can't support itself despite meriting inclusion; in practice, some of the time merges are a compromise between keeping and deleting.
Johnleemk