Jump to content

Wikipedia talk:Categorization

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

This is an old revision of this page, as edited by 98.118.117.123 (talk) at 00:47, 13 January 2011 (→Hidden categories question: new section). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

WikiProject iconManual of Style
WikiProject iconThis page falls within the scope of the Wikipedia:Manual of Style, a collaborative effort focused on enhancing clarity, consistency, and cohesiveness across the Manual of Style (MoS) guidelines by addressing inconsistencies, refining language, and integrating guidance effectively.Manual of StyleWikipedia:WikiProject Manual of StyleTemplate:WikiProject Manual of StyleManual of Style articles
Note icon
This page falls under the contentious topics procedure and is given additional attention, as it closely associated to the English Wikipedia Manual of Style, and the article titles policy. Both areas are subjects of debate.
Contributors are urged to review the awareness criteria carefully and exercise caution when editing.
Note icon
For information on Wikipedia's approach to the establishment of new policies and guidelines, refer to WP:PROPOSAL. Additionally, guidance on how to contribute to the development and revision of Wikipedia policies of Wikipedia's policy and guideline documents is available, offering valuable insights and recommendations.
WikiProject iconCategories
WikiProject iconThis page is within the scope of WikiProject Categories, a collaborative effort to improve the coverage of categories on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.CategoriesWikipedia:WikiProject CategoriesTemplate:WikiProject CategoriesCategories articles

RfC on Sortkey issue

Currently, WP:SORT states "By convention, the first letter of each word in a sort key is capitalized, and other letters are lower case." This has led to the addition of such defaultsorts by bots and other AWB edits. The topic of this RfC is whether this should be a hard rule, implemented on all articles, or a suggestion, only implemented on those articles where it is really an improvement. Fram (talk) 10:45, 2 November 2010 (UTC)[reply]

Discussion

This issue has been discussed and described at the section "Sortkey" on this talkpage. Basically, it boils down to this:

  1. Some pages are incorrectly sorted when they don't have a sortkey added, since sorting is case sensitive and we want it often to be case insensitive. This is the case when you have multiple pages in the same category that start with the same word, but where some have a following word starting with an uppercase, and some with a lowercase.
  2. Some pages are incrrectly sorted when the defaultsort is added to it, and no defautlsort is added to another page in the same category with a name starting with the same word. Before any defaultsort was added, these were sorted correctly.
  3. Adding defaultsorts to all pages would solve issue 2, at the cost of hundreds of thousands of edits which would not improve anything in most cases
  4. Defaultsorts are a problem when a page is moved: without the defaultsort, the title is automatically the sortkey, meaning that a moved page gets sorted under its new name. With a defaultsort, such a page would stil be sorted under its old name. A bot has been suggested to solve this. Fram (talk) 10:55, 2 November 2010 (UTC)[reply]

Examples for the above issues are given in the mentioned "sortkey" section, more examples can be provided if wanted.

My suggested solution is:

  1. Change the rule to a suggestion, that if a page actually gets incorrectly sorted, add a sortkey to the actual category of the page that is a problem, like this: Sort With Upper Case.
  2. Leave all other pages alone: if it isn't broken, don't fix it.
  3. Following this, leave out the addition of such defautlsorts from AWB and other automated or semi-automated edits. Fram (talk) 10:55, 2 November 2010 (UTC)[reply]
Addition (after comments below made it clear that the proposal wasn't sufficiently explicit): defaultsorts for biographical articles should not be affected by this proposal, and should remain: these are in general very useful. Fram (talk) 14:21, 2 November 2010 (UTC)[reply]

Comment on the assumptions above.

3. No one is suggesting adding DEFAULTSORT to all pages. 86% of titles in main-space would not need DEFAULTSORTS for case insensitive sorting (excluding diacritics).
4. Very few pages with a default-sort are moved, less than 100 per day, of those a significant number either

Rich Farmbrough, 13:41, 2 November 2010 (UTC).[reply]
User:Rich Farmbrough/temp111 contains a list for 6.2 days if anyone wishes to analyse it. Rich Farmbrough, 14:06, 2 November 2010 (UTC).[reply]
E.g. Kidnapping of Jalal Sharafi, clearly incorrect defaultsort after move. Black-lored Parrot had a defaultsort added through AWB: while the page move hasn't really changed this, it is one of three black-x birds sorted out of order at Category:Birds of Indonesia, where 25 others are not sorted by this rule. Not having a defaultsort would have been better in this case as well. The same goes for different other moved bird articles from the list as well. Ulnar collateral ligament of thumb was sorted before Ulnar carpal collateral ligament before the move, and still is afterwards.
Basically, looking through that move log, and ignoring articles where this RfC wouldn't make a difference (mainly articles on persons), I see multiple articles where not having a defaultsort would be better, and none where the defaultsort actually made an improvement. Fram (talk) 14:21, 2 November 2010 (UTC)[reply]
Oh, and I have reverted your AWB change here, since you changed a defaultsort that editors had added specifically to be different from the article title, to one matching the article title, thereby ruining the prupose of their defaultsort. Please, if the pre-move defaultsort didn't match the article title, there is no need to make it so post-move, as one may think thta the difference was done deliberately... Fram (talk) 14:37, 2 November 2010 (UTC)[reply]
I analysed over 100 and only 4 had defaultsorts that would have been put on by AWB before the move, and were wrong afterwards. Rich Farmbrough, 01:02, 4 November 2010 (UTC).[reply]
And how many had defaultsorts that would have been put on by AWB, and were not wrong afterwards? Not 96%, obviously, as these incldue many biographies... And of those that would have been put on by AWB, how many made an actual positive difference to the sorting of the article in even one category? More than those 4%?Fram (talk) 12:58, 4 November 2010 (UTC)[reply]
Not all pages, but many hundreds of thousands of pages. 14% of the titles still means close to 500,000 pages.
About your page move examples: thanks for providing such a nice example. The Category:London Film Critics Circle Awards lists four specific awards. If none of them had a defautlsort (my preferred state), all four would sort correctly. Currently, due to changes you made to them with AWB[1], they have a defaultsort which hasn't improved any sorts. However, if someone would (quite understandable) change the defaultsort for London Film Critics Circle Award for Actress of the Year from the current, pre-move "DEFAULTSORT:London Film Critics Circle Award For Best Actress" to the correct "DEFAULTSORT:London Film Critics Circle Award For Actress Of The Year", it would no longer sort correctly. So this is, contrary to your claims, a series of articles where no defaultsort would have been better; they would all have sorted correctly, and no changes were necessary after the move either.
The first one you provide, "Carrier sense" etcetera, is currently, thanks to an unnecessary defaultsort, incorrectly sorted at Category:Channel access methods: the two articles that received a defaultsort through Smackbot edits are now sorted before the main article, which doesn't have a defaultsort. If Smackbot hadn't added the two defaultsorts, these articles would sort correctly, thanks to the page move, which now indeed had no effect on the sorting.
Your examples three and four are about a person, where there is no objection whatsoever to adding a defaultsort in nearly all cases.
As was discussed in the previous section on this: can you provide some insight into why this should be a hard rule for every article that matches this title description, and not a soft rule to be implemented only when it is really improving things? Fram (talk) 14:08, 2 November 2010 (UTC)[reply]
  • If Template:Bug were fulfilled, a great deal of of the DEFAULTSORT would become unnecessary. We should try and have the bug fixed and eliminate the need for these edits. –xenotalk 13:59, 2 November 2010 (UTC)[reply]
  • I see no reason why every page should not have a sortkey by default. Debresser (talk) 14:29, 2 November 2010 (UTC)[reply]
    • What would be the benefits of this? Disadvantages plenty: first a run through all 500000 pages or so that would need one with this rule: then new bots or new maintenance (backlog) cats, with "pages missing a defaultsort" and so on, since new pages will not automatically get a defaultsort, making it necessary to go through these pages every day: all this for what prupose exactly? What is the ratio of pages improved by this, vs. pages needlessly edited by this? Fram (talk) 14:37, 2 November 2010 (UTC)[reply]
  • Going through every category to find pages that are sorted incorrectly would be harder than using a bot to go through every page and add an automatic sort key. Problem 4 isn't a problem with the sort keys, it's a problem with editors. Editors should just change the sort key when the page is moved. Simple. McLerristarr | Mclay1 08:10, 5 November 2010 (UTC)[reply]
  • Yes it's harder but I have code to do it on dumps already. This might help those who don't want to see a single un-necessary edit. Rich Farmbrough, 08:44, 12 November 2010 (UTC).[reply]
  • I propose that we add DEFAULTSORT to every single page. 1/3 of Wikipedia pages are biographical and a big percentage uses special characters. Adding DEFAULTOSRT to the rest won't be a big problem as soon as we have a good estimate of how many pages we are talking about. After we finish adding DEFAULTSORT to all pages we can just have a bot to check if pages moved daily have the correct sortkey. -- Magioladitis (talk) 12:58, 5 November 2010 (UTC)[reply]
    • ...and a second bot to add defaultsort to every new page (and every page moved from another namespace to the main namespace) as well (not intended as criticism of your opinion, just an indication of what is needed if we indeed go this way instead of my way (which is rather lonely so far, sob sob ;-)).
There are two BRFA's in already from me and Rjwilmsi, to deal with moves that need changing. Adding new pages would be a cinch I think. Rich Farmbrough, 08:44, 12 November 2010 (UTC).[reply]
Two quick points:
  • The last point under Sort Keys says, "Default sort keys are sometimes defined even where they do not seem necessary—when they are the same as the page name, for example—in order to prevent other editors or automated tools from trying to infer a different default." I think this should be changed to "Default sort keys should be defined even where they do not seem necessary—when they are the same as the page name, for example—in order to prevent other editors or automated tools from trying to infer a different default." A bot could just as easily sort Pink Floyd as "Floyd, Pink" as a bot sorted Ptolemy I Soter as "Soter, Ptolemy I". (I kid you not. It did happen.)
  • The only pages that absolutely do not need a sort value are those with one-word titles. JimCubb (talk) 17:12, 10 January 2011 (UTC)[reply]

Moving forward

Seeing this has stalled for 2 weeks now and two BRFAs by an editor involved are waiting approval, I will attempt to at least partially move this issue forward.

So: Should DEFAULTSORT be added to articles where it does not impact the actual sorting of the article? Hypothetical example: add DEFAULTSORT to ReD as "Red" when it is already sorted correctly between Rack and Ruth.

  • Oppose. Don't fix what's not broken. A bot/script/AWB can (be modified to) determine when adding a sortkey will actually change the sorting order. —  HELLKNOWZ  ▎TALK 13:18, 28 November 2010 (UTC)[reply]
  • Yes, it should. At any given time, it is unknown what entries will be in each of a page's categories or in which categories an editor will include a page; excluding a defaultsorts in cases where it currently does not matter creates inherent errors when more entries are added to categories. While the sorting is still sorted based on ASCII (or does not account for case or accented characters), the defaultsort is useful, especially on pages in mixed case like ReD. —Ost (talk) 20:02, 2 December 2010 (UTC)[reply]
  • Oppose. It is difficult enough to figure out sorting problems being created by the stupid bots that are adding default sort incorrectly. Don't use DEFAULTSORT unless a human actually determines that it is needed. Like HELLKNOWZ said, Don't fix what's not broken. Vegaswikian (talk) 20:17, 2 December 2010 (UTC)[reply]
  • Comment While I don't support a bot-run to add DEFAULTSORT to all articles, I do not oppose it either (and would be happy to do it if consensus requested it). Nonetheless this is a slightly false dichotomy. Given our current standards there are more than two actions we could take:
  1. DEFAULTSORT to all articles
  2. DEFAULTSORT to articles which could possibly be mis-sorted
  3. DEFAULTSORT to articles which are mis-sorted

And within this

  1. Do it regardless
  2. Do it only if making another edit.

(Also there are a few articles which benefit from a DEFAULTSORT equivalent to the article title, for relatively obscure reasons.)

Now I have always gone with do 3. regardless, and do 2 if you are editing the article anyway, as the behaviour we should aim for.

Further I would suggest we can have no issue with people who don't add a DEFAULTSORT - we might point it out to them if they are creating many articles, or even editing lots of them (some stub-sorters do this as they go I believe) but we should not require them to do it. Moreover I have no problem with anyone adding them more zealously. There are perfectly good arguments for doing so. Rich Farmbrough, 15:25, 19 December 2010 (UTC).[reply]

Not knowing about this thread, I asked a few similar questions at Help talk:Category#Defaultsort. In particular, why edits like this one? — Pt (T) 16:01, 2 January 2011 (UTC)[reply]

What is the reason for the convention that the first letter of each word in a sort key is capitalized? Given that only the first word and proper names are capitalized in article titles according to WP:LOWERCASE, I find the sort-key convention counter-intuitive and surprising and I guess it has been discussed somewhere. Nurg (talk) 23:03, 2 January 2011 (UTC)[reply]

Probably because multi-word proper names (each word capitalized) were found or felt to be more common as article titles than multi-word phrases with only the first word capitalized (so adopting that as the convention would put more articles automatically in the right place).--Kotniski (talk) 09:32, 3 January 2011 (UTC)[reply]
And it seems this is right: I have counted all the multi-word article titles (including redirects) based on the last database dump from October. And 1675023 articles use only lowercase, 3707972 articles use only uppercase and 3317117 don't fit either (e.g. they had both lower and upper case, contain numbers or have disambiguation part). I considered words to be delimited by spaces only. Svick (talk) 21:52, 8 January 2011 (UTC)[reply]

Category:Eponymous categories

Please check it out. I am used to the weirdness of Wikipedia, but the stupidity of this boggles the mind. Never mind the dubious use of the word "eponymous", this appears to be some sort of "meta" or "shadow" category system: since every category is necessarily named after something, hey, we can categorize each category as a category named after something. Category:Religious texts? .... Category:Categories named after religious texts. And so on. Wait, once we have duplicated our category system, why not introduce categories that categorize categories created to categorize categories? "Category:Categories categorizing categories named after religious text". Neat, isn't it?

I am not sure who is the genius we have to thank for this ... novelty, but the, ahem, general approach reminds me of the "outlines" crew. Basically, "the encyclopedia is written, and all our finished articles are so perfect we are now bored, so now let's duplicate it into a mirror-encyclopedia", again and again. --dab (𒁳) 08:37, 10 December 2010 (UTC)[reply]

I'm not sure I can see the full ramifications of what you suggest, but I must admit I have not personally been able to see any actual utility in having this category scheme. __meco (talk) 09:49, 10 December 2010 (UTC)[reply]
Is there any actual problem or proposal here? (Confusion sometimes results from the fact that we're not allowed to make a category a "member" of another category - they have to go in as "subcategories", even if that's not logically what we mean - but this is a problem with the software, and I think the developers long ago gave up on making any improvements to the categorization function in MediaWiki.)--Kotniski (talk) 10:40, 10 December 2010 (UTC)[reply]
I would suppose that the problem is "creation of unnecessary, internal categories without any noticeable benefit" and that the proposal is "get rid of them". If the problem is correctly identified (and I have seen no indication that it isn't), then the proposal is logical and good. Fram (talk) 10:56, 10 December 2010 (UTC)[reply]
The benefits of eponymous categories were discussed at Wikipedia:Categories_for_discussion/Log/2010_August_12#Category:Eponymous_categories. Epbr123 (talk) 11:03, 10 December 2010 (UTC)[reply]
Basically I think they're just containers, on the basis that everything ought to be categorizable somewhere (except the category at the absolutely highest level, which is called Category:Contents or something). They hardly get in the way, and might occasionally be used by someone (like if you're wondering how to organize an eponymous category for a religious text, you might want to see how it's been done for other such entities).--Kotniski (talk) 11:27, 10 December 2010 (UTC)[reply]
(ec)All I see there are a number of people having trouble grasping the concept that a category can be a subcat of another category, instead of an article being in that category. That the category Kylie Minogue is a subcat of the category Australian singers seems perfectly logical to me, and thus I see no reason to have these eponymous categories, and would support their deletion. If I want to see articles related to Belgian people, I go the the cat Belgian people, where I expect to find all articles on Belgian people and all categories with articles directly reating to Belgian people, like Category:Georges Simenon. However, that cat is only placed in the Category:Categories named after writers... Fram (talk) 11:29, 10 December 2010 (UTC)[reply]
If we want to have a hierarchical category system, then that's perfectly logical (it isn't logical to make Category:Georges Simenon a subcategory of Belgian people, since not everything connected with Simenon will be a Belgian person). If we base categories always on things being related to other things (and not being a certain type of thing), then we would end up with a vague cloud (since everything is related to everything else via a few steps).--Kotniski (talk) 11:36, 10 December 2010 (UTC)[reply]
> That Category:Kylie Minogue is a subcat of the Category:Australian singers seems perfectly logical to me
That's because you're not distinguishing between topics and sets. Category:Australian singers is a set, so it's supposed to include only articles about singers. Category:Kylie Minogue is a topic, so it can contain anything significantly related to the singer it's named after. —Coder Dan (talk) 17:48, 10 December 2010 (UTC)[reply]
I'm indeed not making that distinction, as I am not interested in such distinctions. Categories are intended to be an easy and logical search and classification system, where are related articles can be found starting from an obvious top. Category:Australian singers should contain all articles about and directly relating to Australian singers, so Kylie Minogue but also Kylies discography, tours, and individual records of her (as they are music by Australian singers, Tours by ustralian singers, ...: all logical children of the parent cat). On the other hand, het tours are not directly related to the topic "1960 births" (or whicheber year she's from): I expect the article on Kylie to be in a birthyear category, but not the category on Kylie: I do expect the category Kylie to be in the Australian singers cat though, as all articles in that cat are direct, on topic descendants of that category. Fram (talk) 20:54, 10 December 2010 (UTC)[reply]
Well those distinctions exist in the Wikipedia category system, so please don't accuse other editors of "having trouble grasping the concept" of subcategories just because they get their definitions from dictionaries and you get yours from TV game shows. As I said in another post, there should be a two-tier hierarchy that includes both inclusive (topic/TV) and exclusive (set/dictionary) categories. —Coder Dan (talk) 21:23, 10 December 2010 (UTC)[reply]

the point is that these categories should have been deleted on sight as obvious nonsense. Instead this was turned into a wikidrama, with emotional defenses of their utility that frankly are too surreal for me to follow. Please delete them.

Category:Georges Simenon can be a subcategory of Category:Belgian people, plus any number of relevant categories. There is no rationale for a Category:Categories named after writers unless we decide that categories are not just for the categorization of articles, but also for the categorization of categories, and that the categorization of categories for some reason should be a system entirely distinct from the categorization of articles. At this time my mind breaks down over the tangible thickness of this. As Fram points out, all this seems to be about is that some people fail to grasp the concept of "hierarchical". --dab (𒁳) 12:31, 10 December 2010 (UTC)[reply]

I'm not sure what you mean now - if you think it makes sense to make Category:Georges Simenon a subcategory of Category:Belgian people, then we sacrifice any kind of sensible logical hierarchy (this does happen, of course, with countries for example, and it leads to all articles on people being part of the hierarchies for Geography and Chronology, since all people were born at some time and some place; if we do the same thing with eponymous categories for people I suppose it just leads to more absuridities of the same type - all paintings would be under Geography because they were painted by someone who was born in some place - except they wouldn't all be, because not all painters have eponymous categories containing their paintings... do you see what I mean?)--Kotniski (talk) 14:02, 10 December 2010 (UTC)[reply]
A practical advantage of keeping eponymous categories and categories such as Category:Belgian people seperate is that it allows bots and AWB users to compile lists of people articles, enabling reports such as Wikipedia:Database reports/Potential biographies of living people (4) to be created without without a large number of false positives. If you want to make it easier for readers to find Category:Georges Simenon, a "see also" link could be added to Category:Belgian writers. Epbr123 (talk) 15:56, 10 December 2010 (UTC)[reply]

Personally, I've never been a fan of "Category:Categories named after FOO" or "Category:Eponymous categories", or anything of the sort. No-one has been able to convince me that these are needed. --Kbdank71 16:43, 10 December 2010 (UTC)[reply]

I was surprised to see Category:Wood and Category:Oil palm in Category:Eponymous categories but they were added very deliberately by an admin who removed all other categories.[2][3] Is this within the intended use of the category? PrimeHunter (talk) 17:04, 10 December 2010 (UTC)[reply]

That just completely screws up navigation. You shouldn't have to jump back and forth arbitrarily between articles and categories to find other categories or articles. The only practical reason I can see for the different treatment of "eponymous categories" (aka topical categories) is the automated list generation cited above by Epbr123. But there are much less convoluted and navigation-hindering ways to do that. Make Category:Eponymous categories (or as I'd prefer it named, Category:Topical categories) a hidden category and have any search exclude subcategory entries with that tag. postdlf (talk) 17:12, 10 December 2010 (UTC)[reply]

> I'd prefer it named, Category:Topical categories
I would just call it Category:Contents. Topics naturally include sets as subcategories, so Category:Topical categories effectively includes all articles on Wikipedia. —Coder Dan (talk) 18:02, 10 December 2010 (UTC)[reply]

The problem is that the logical argument for eponymous categories is actually illogical. If "Category:John Smith" was a sub-category of "Category:English people", then the article "John Smith's pet gorilla" in "Category:John Smith" would be, by extension, in "Category:English people", which is wrong. But similarly, if "Category:John Smith" was in "Category:Categories named after people", then "John Smith's pet gorilla" would be, by extension, in "Category:Categories named after people", which is equally wrong. So really, neither system works. All articles are also in, by extension, Category:Main topic classifications. So either we ignore the whole "categorised by extension" problem and do away with eponymous categories or we need to think up a new system. McLerristarr | Mclay1 17:16, 10 December 2010 (UTC)[reply]

Good points. That whole way of analyzing category nesting has always struck me as wrong-headed and strangely ideological, based on some view of classification purity at the expense of every pragmatic consideration. You've pointed out well how that "purity" breaks down even under the supposed solution. postdlf (talk) 17:19, 10 December 2010 (UTC)[reply]

The stupidity is in the naming system, not the categories themselves. Eponymous categories are just topic categories that happen to have eponymous articles. Category:Georges Simenon can't be a subcategory of Category:Belgian people, because Category:Georges Simenon contains everything that has any notable connection with a particular person, while Category:Belgian people contains only articles about the people themselves. If Category:Georges Simenon and Category:Belgian people were called Topic:Georges Simenon and Set:Belgian people, we wouldn't have to waste so much time with these stupid arguments every few months. —Coder Dan (talk) 17:38, 10 December 2010 (UTC)[reply]

The division isn't so clear: Category:Belgian people also (logically) includes List of Belgian people, which is itself not a person article. And it also (logically) contains Category:Belgian diaspora, which in turn contains Belgian Brazilian, Luxembourg Brazilian... Categories are just groups of articles that pertain to the category's name. The problem instead is in the completely unjustified expectation that Category:Wolves only contain types of wolves, rather than also articles about wolves such as wolf-baiting, History of wolves in Yellowstone, Category:Deaths due to wolf attacks, etc. Articles about Georges Simenon are articles about a Belgian person. Once again, to do it your way only accomplishes the artificial separation of related articles, with nothing gained. You're not the only editor to have your view obviously, but other than the automated listing issue (which can be solved in other ways), I have seen no concrete problem described, just some kind of intellectual distaste. No reader will ever say "the inclusion of Category:Georges Simenon in Category:Belgian people irrevocably confused me" or "I wrote a report on wolves and got an F because I said "wolf-baiting" was a type of wolf." postdlf (talk) 17:57, 10 December 2010 (UTC)[reply]
> The division isn't so clear
Nothing is perfectly clear in the real world, but that doesn't mean we should abandon the hierarchical model altogether. Some readers want to read everything about some particular topic, others only want to read about the members of a set. The distinction between things and lists of things is fairly trivial.
> Categories are just groups of articles that pertain to the category's name
No they're not. Categories are classes of things that share some property. Category members don't just pertain to the category's name, they're defined by it. Categories as general associations is a misconception that's fostered by television game shows.
> No reader will say "inclusion of Category:Georges Simenon in Category:Belgian people confused me"
The problem isn't confusion, it's noise. Some readers don't want to wade though a heap of tangentially related articles just to read about a group of individuals.
—Coder Dan (talk) 18:25, 10 December 2010 (UTC)[reply]
Re: pertaining vs. defining, I'm not talking about putting agriculture in Category:Kansas here. This is semantics. Wolf-baiting is defined by being about wolves, the contents of Category:Georges Simenon are defined by being about a particular Belgian person. I would expect to find both by navigating through those respective category systems.

Re: noise, subcategories are already visibly separated from articles in the category page. Pipe sorting of subcategories can separate those that are further subcontainers (i.e., see how Category:European countries is sorted) from those that are subtopics. Also, as with any category's members, the Belgian people who have their own subcategories are likely to be the more important ones, given that they have multiple articles about them and their works, so arguably those should stand out. If I'm trying to find the article on Georges Simenon in Category:Belgian writers in French (because I can't remember how to spell his name, just that it starts with an S) not only would I see that there is an article on him, but a whole category of articles about him if Category:Georges Simenon were actually categorized as the article is. If I'm trying to find Jean d'Osta in Category:Belgian writers in French, if Category:Georges Simenon were also placed in that category how would it hinder that search? postdlf (talk) 18:47, 10 December 2010 (UTC)[reply]

> This is semantics.
This entire conversation is about semantics, and my Merriam-Webster's seems to agree with my definition of "category", so you should stop throwing the word around as if anyone who disagrees with you is automatically wrong.
> I would expect to find both by navigating through those respective category systems.
I know what you expect. As Epbr123 pointed out earlier, this was already discussed ad nauseum on Wikipedia:Categories_for_discussion/Log/2010_August_12#Category:Eponymous_categories. I would expect to find sets of things without having to wade through piles of tangentially related articles, so there should really be a two-tier hierarchy, with a Topic namespace that includes eponymous categories and a Set namespace that excludes them. —Coder Dan (talk) 21:06, 10 December 2010 (UTC)[reply]

Playing with an example

I'm looking at Category:Horses. Now it belongs, through a long-winded chain of subcategories, to Category:Nature, Category:Life, Category:Agriculture, and Category:Science, all of which are fundamentals. Why science? Because it's in Category:Animals, which traces through zoology and so forth. When one moves up a level, the descendants split out quite a bit further: there are categories for types of horses and breeds of horses, which I suppose could fit within the nature and life and generally agriculture supercategories, and then there categories like Category:Horse tack which descend into Category:Technology ultimately, and then there is Category:Famous horses. So we have one subcategory tree of horses and another of horse-related things.

The "eponymous" rule implies that any category with a parent article about a specific thing or class thereof should not have any parent categories; therefore (for instance) there should be no category tree of sciences, but there might be a subcategory of Category:Science which contains all the articles on specific fields. Likewise, Category:Horses would have no parent, but there could be categories under it, all of which would be intersections with other category trees or diffusion categories which grouped the intersections together. And therefore the horse breed category structure would also descend from some Category:Breeds, and Category:Famous horses would descend from some Category:Famous individuals somehow.

I think we could do this, but I think people would find it very confusing, because to trace backwards one would constantly be switching back and forth between categories and articles. It would mean that everything except "X by Y" categories and their children would be a category root. Mangoe (talk) 18:36, 10 December 2010 (UTC)[reply]

I see people arguing against categorizing the cat "The Beatles" in the cat "British Pop Bands", because that would mean that "Revolver" would become indirectly a British Pop Band... Using that logic, and putting "The Beatles" in "Eponymous band categories", doesn't that mean that "Revolver" is also an "eponymous band category"? It is just as much a grandchild of that cat as it was of British Pop Bands. Using this argument, we shouldn't have topical categories, only set categories, which would npot only exclude the eponymous cats, but also all cats inside them... Fram (talk) 20:58, 10 December 2010 (UTC)[reply]
There doesn't appear to be a Category:Revolver (album), and the songs on it (e.g. Yellow Submarine) are all in either property or property intersection categories, but none of them is categorized as being on that album. However, Category:The_Beatles is in Category:1960s music groups, which strictly speaking is wrong according to the strict version of this theory: the only "valid" parent is Category:Categories named after musical groups, which is a direct child of eponymy. Mangoe (talk) 21:34, 10 December 2010 (UTC)[reply]
Category:The Beatles being in Category:1960s music groups is my fault. I added it a while ago before I properly thought about the eponymous categories thing. I have fixed it now. Category:Horses fits perfectly well where it is. There may be some problems but that's due to miscategorisation. Categories such as Category:The Beatles could also have parent categories such as Category:Rock music. I have no idea why this is not already the case. McLerristarr | Mclay1 09:16, 11 December 2010 (UTC)[reply]
By the way, Fram was talking about the article Revolver (album), not a category. McLerristarr | Mclay1 09:17, 11 December 2010 (UTC)[reply]
> doesn't that mean that "Revolver" is also an "eponymous band category"?
Yes, which is why it's stupid to put "categories" in the names of categories. Eponymous categories are really just a way for naïve editors to create a topic hierarchy on top of the existing set hierarchy. —Coder Dan (talk) 21:39, 10 December 2010 (UTC)[reply]
No, Revolver (album) can in no way be interpreted as an eponymous band category. McLerristarr | Mclay1 09:18, 11 December 2010 (UTC)[reply]
Okay, maybe the people who create those categories really are so clueless that they believe the presence of an eponymous article is important. But they also have a legitimate purpose: All of those categories are topics, so maybe it would be good to organize them separately from set categories. —Coder Dan (talk) 11:02, 11 December 2010 (UTC)[reply]

Proposal to change the eponymous category system

To reiterate my earlier point with a real example, the current situation is this: Category:The Beatles, for example, is in Category:Categories named after musical groups rather than in categories like Category:1960s music groups. The argument for is that a topic category cannot be a subcategory of a list category because if Category:The Beatles were a subcategory of Category:1960s music groups, then Paul is dead would be, by extension, in Category:1960s music groups despite not being an English pop band. However, under the current scheme, Paul is dead is, by extension, in Category:Categories named after musical groups despite not being a category at all. Neither system works. I propose that we get rid of Category:Eponymous categories and its related subcategories and use a new categorisation system that still keeps topic categories out of list categories. The eponymous topic categories can be categorised into other topic categories. For example, Category:The Beatles could be categorised in categories such Category:1960s in music and Category:Rock music.

...Which still puts articles like Paul McCartney 2004 Summer Tour in Category:1960s in music. People's attempts to get a perfectly logical categorization system are a wild dream that can not be achieved on Wikipedia, and isn't necessary either. Fram (talk) 08:24, 11 December 2010 (UTC)[reply]
Well, Category:Paul McCartney shouldn't be a subcategory of The Beatles really. It should be the other way around. McLerristarr | Mclay1 08:56, 11 December 2010 (UTC)[reply]
I've swapped the categories around so that Category:The Beatles is a subcategory of The Beatles' categories and not the other way around. Categories of musicians seems to be only categorised under Category:Categories named after musicians. Under my proposal, this would be changed to relevant genre categories. Musicians and groups spanning multiple decades probably shouldn't be categorised in decade categories to avoid problems such as Paul McCartney 2004 Summer Tour being categorised in Category:1960s in music. McLerristarr | Mclay1 09:07, 11 December 2010 (UTC)[reply]

There may always be some small things in the categorisation system that don't quite fit, but the current system is worse. Most problems with categorisation are more to do with miscategorisation rather than problems with the system itself. McLerristarr | Mclay1 08:59, 11 December 2010 (UTC)[reply]

Putting Category:The Beatles in Category:1960s in music doesn't really solve the problem. What we really need is something like Topic:1960s music groups or Category:1960s music groups (topic). Then you can put anything you want into it without interfering with Category:1960s music groups. We need an entire hierarchy of topic categories that sits on top of the existing set-oriented hierarchy. —Coder Dan (talk) 11:16, 11 December 2010 (UTC)[reply]

One thing that is really needed is a way to include a category as a member of another category, rather than a sub-category. But the main thing that is needed is understanding and attention. For example "Paris Metro" was a sub-category of "French railway stations" (I may not have the names exactly right) effectively making certain items of rolling stock into stations. It takes a only a few minutes to correct this sort of thing. Similarly with eponymous town categories, which end putting their entire contents in "Populated places in..." worse still, if they include the eponymous categories of their "notable inhabitants" then a place in India can be in a sub cat of "populated places in New York". The basic options are:
  1. Dump eponymous categories (such as Category:The Beatles) altogether
  2. Fix up on an on-going basis incorrect sub-cats by moving them into the correct parent categories, possibly using explicit titles (such as Category:Articles related to the Beatles).
  3. Dump the enormous benefits of having subcategories that actually contain what the parent category implies they will
Rich Farmbrough, 12:33, 11 December 2010 (UTC).[reply]
Yes, I think the second option would be the most helpful, though it has to be admitted it would be quite a task, since there are very many cases of categories being made subcategories on the basis of "but it's named after a member of the category" (and many editors who don't understand - or don't accept - why that might be considered an invalid reason). For me, though, the most unhelpful practice is when people take articles out of a category on the grounds that their eponymous categories are already there - this leads to incomplete alphabetical lists within the category, and (often randomly) incomplete lists of categories on the article pages. --Kotniski (talk) 12:46, 11 December 2010 (UTC)[reply]
As one of the editors who doesn't "accept" this approach, I will agree with you on the last point: articles should never only be placed in an eponymous category for just the reasons you have given. postdlf (talk) 13:32, 11 December 2010 (UTC)[reply]
> include a category as a member of another category
That's not really necessary. If you're looking for someplace to put Category:Articles related to the Beatles, all you need is Category:Articles related to 1960s music groups. The whole point of putting "Articles related to" in a category name is to label the category as a topic. Category:The Beatles is already a topic category, so all your second option does is rename it. The only thing that's really missing is a larger topic category to contain it as a subcategory. There's no need to make the category itself a member. —Coder Dan (talk) 15:40, 11 December 2010 (UTC)[reply]
In reply to Coder Dan, I don't see why putting Category:The Beatles in Category:1960s in music doesn't solve the problem. Category:1960s in music basically is a Topic:1960s music groups.
Category:1960s in music is a broader topic than Topic:1960s music groups. Category:1960s in music contains everything connected with music and the 1960s, while Topic:1960s music groups has the additional restriction of being related to groups. It's a legitimate move to make, but it doesn't solve the problem that some people will still want to put Category:The Beatles in Category:1960s music groups. —Coder Dan (talk) 15:51, 11 December 2010 (UTC)[reply]
In reply to Rich, dumping all the eponymous categories would make the categorisation system far worse than it already is. Category:The Beatles neatly groups all the relevant Beatles articles and I see no other way of grouping them.
I still haven't really heard any reasons why people disagree with just putting the topic categories into other topic categories. Most categories are already like this e.g. Category:Horses. Beatles → Rock → Music → Arts. Seems simple to me. McLerristarr | Mclay1 15:31, 11 December 2010 (UTC)[reply]
Most categories are not like that. Countries and animal species are exceptions to the rule. —Coder Dan (talk) 15:56, 11 December 2010 (UTC)[reply]
OK, well, I didn't really count them and I'm assuming that neither did you. Country and animal categories are quite a lot in number but that's beside the point. My point is most categories can and should be categorised like that. McLerristarr | Mclay1 15:59, 11 December 2010 (UTC)[reply]
> I'm assuming that neither did you
You can assume anything you want, but it won't make you any smarter. The guideline explicitly designates plural-named categories as sets, and crowding those categories with extraneous material would be very annoying to people who only want to read articles about the things themselves. My point is that there should be two sets of categories, one categorized your way (as topics) and one categorized the current way (as sets). —Coder Dan (talk) 16:11, 11 December 2010 (UTC)[reply]
I agree with you! I'm not suggesting we mix the two together. We only differ in opinion in that what you are suggesting is to make an entirely new namespace for topics. Perhaps it may be best to get rid of the eponymous category system first, then think about such a monumental move as you are suggesting. But the hard fact is nothing will ever stop people wanting to cagorise Category:The Beatles in Category:1960s music groups. But that's a problem with editors, not the system. Luckily, very few inexperienced editors edit categories. McLerristarr | Mclay1 16:28, 11 December 2010 (UTC)[reply]
You said "most categories can and should be categorised [as topics]." When you say "Most categories should be topics", it sounds to me like you want to replace sets with topics, whereas I want to add the topics. As far as I can tell, that's our main disagreement. I want to retain a full hierarchy of set categories, including exclusive categories for current exceptions, such as countries and biological species. My main points are that (a) sets and topics should co-exist in some logical way and (b) they should be clearly labeled to avoid the confusion typified by this discussion. Namespaces are a way of making my point, but they're not likely to happen in the short run. I don't think many people will insist on putting Category:The Beatles in Category:1960s music groups if it's already in Category:Articles related to 1960s music groups or Category:1960s music groups (topic) or whatever. —Coder Dan (talk) 17:32, 11 December 2010 (UTC)[reply]
No, that's not what I meant. Set categories should be kept as is. I meant most topic categories can be categorised as subcategories of other topic categories rather than in an eponymous category category. I understand your point about wanting to make a Category:1960s music groups (topic). If every large set category had a corresponding topic category, it would be problem solved. Category:Presidents of the United States has Category:Presidency of the United States. At the moment, the closest topic category to Category:1960s music groups is Category:1960s in music. I really think that if we can just get rid of the eponymous category categories first, then we can start trying to create more topic categories. McLerristarr | Mclay1 04:25, 12 December 2010 (UTC)[reply]
> get rid of eponymous category categories first, then create topic categories
Why not do both together? Each "Categories named after Xs" category should be renamed to something like "Articles related to Xs" or "Xs (topic)" and moved to some reasonable place. —Coder Dan (talk) 10:55, 12 December 2010 (UTC)[reply]

OK, well until we can think of a better way of doing it, I guess we could do that. But some of these could easily be incorporated into existing categories. For example, Category:Disney's Aladdin, could be put in Category:Disney. Admittedly, it would make some topic categories, such as Category:Disney, a lot bigger but that's not too much of a problem, is it? However, quite a few set categories contain articles and subcategories. For example, Category:European countries contains United Kingdom and Category:United Kingdom. Is this a better way of doing it? Either way, all categories should be consistent, we need to keep them all following the same guidelines. McLerristarr | Mclay1 13:28, 12 December 2010 (UTC)[reply]
I think what we to do is rename or delete categories with "category" or "categories" in the name (with obvious exceptions). "Categories by X" can become "Topics by X". McLerristarr | Mclay1 13:51, 12 December 2010 (UTC)[reply]

I can't tell if you're being facetious or not, in saying we could just arbitrarily decide to get rid of many categories that sort a substantial number of articles, just because we can. Let me know when you post those CFDs so I can see how those arguments proceed. postdlf (talk) 14:48, 12 December 2010 (UTC)[reply]
> Category:Disney's Aladdin, could be put in Category:Disney
Don't worry about that. We're not talking about Category:Disney's Aladdin, we're talking about Category:Categories named after films and its siblings.
> Category:European countries contains Category:United Kingdom
That needs to be fixed. One reason these arguments go on so long is because people who favor categories as topics can point to exceptions such as countries and animals as supportive examples.
> "Categories by X"
If you want to do that, then start a new section on categories of categories or "Categories by" categories. This section is about "Categories named after" categories.
—Coder Dan (talk) 15:09, 12 December 2010 (UTC)[reply]
Yes, perhaps my second comment was a little off-topic. My point about categories like Category:Disney's Aladdin is that if they were dispersed into relevant topic categories, categories like Category:Categories named after films would be empty and no longer needed, thus saving us from needing to redefine or rename them. McLerristarr | Mclay1 15:18, 12 December 2010 (UTC)[reply]
> if [subcategories] were dispersed ... [eponymous-category categories] would [not be] needed
That's not true. It's still useful to have collections of articles related to the topics defined by these categories. As I said earlier, there's nothing intrinsically wrong with these categories except the requirement of having an eponymous article. They just need to be renamed and recategorized, and that's easy to do if we can clear up the confusion about them. —Coder Dan (talk) 15:57, 12 December 2010 (UTC)[reply]
Ah, that's a very good point and have no idea why I did not think of that. It is useful to group these categories together. OK, I think we should formulate a CFD proposal. McLerristarr | Mclay1 16:04, 12 December 2010 (UTC)[reply]
We should agree on some kind of (re)naming convention. So far we have "Articles related to Xs" and "Xs (topic)". —Coder Dan (talk) 17:49, 12 December 2010 (UTC)[reply]
I think "related to" implies even looser inclusion criteria than my "pertaining" choice of words above that you didn't like. Rather than adding such self-referential and awkward phrases to category names, why not handle it through template tagging, as Mclay has suggested below? That way they could also be added to hidden categories, which would aid with the automated list generation issue. postdlf (talk) 18:03, 12 December 2010 (UTC)[reply]
> my "pertaining" choice of words above that you didn't like
That has nothing to do with what we're talking about. You used the word in a definition, not a category name. Category:Articles related to music groups is a group of articles that are defined as being related to music groups. Category:Music groups is a group of articles that are defined as being about the music groups themselves. In both cases, category members don't just pertain to the category's name, they're defined by it. I don't know what you think is awkward or self-referential, but naming categories has nothing to do with templates or hidden categories. We just need a scheme for naming a set category and a topic category about the same kind of thing. —Coder Dan (talk) 19:13, 12 December 2010 (UTC)[reply]
Perhaps Category:Music group topics or Category:Topics by music group? I like the Category:Music groups (topic) disambiguation. We could rename Category:Music groups to Category:Music groups (set) if it would make things clearer. McLerristarr | Mclay1 03:45, 13 December 2010 (UTC)[reply]
"Music group topics" is okay, but "By" has a special meaning in Wikipedia category lingo. Category:Xs by Y is a metacategory that divides the set Category:Xs according to property Y. I guess postdlf might like "Articles pertaining to Xs". "(set)" makes sense technically, but I think it would be enough to just strengthen the language in the guideline that defines Category:Xs as a set. —Coder Dan (talk) 15:41, 13 December 2010 (UTC)[reply]

If the underlying "problem" is that "topic" subcategories are somehow getting in the way of "set" subcategories, then why not have a developer-based solution that would somehow separate those contents within a category? Any time a category and an article have the same name (as should be the case in most "eponymous" categories; for those that don't match, there would be some manual way of tagging them differently), those could display as "member categories" separate from "subcategories", and then automated listmaking could be set to ignore the "member category" contents. That would solve the problems identified above (to the extent they are concrete enough to make sense to me) and still preserve the unity of the category system for grouping and navigating related articles. postdlf (talk) 14:48, 12 December 2010 (UTC)[reply]

I'm not entirely sure if I understand what you just wrote. Would I be correct in saying that you mean keep topic categories and set categories listed separately in a category rather than keeping them all together in the subcategory section? McLerristarr | Mclay1 15:07, 12 December 2010 (UTC)[reply]
Basically, though it wouldn't function on a "topic"/"set" dichotomy, except to the extent that "topic" = "subcategory defined by an article within the parent category". Thus, as France and United Kingdom are articles in Category:European countries, Category:France and Category:United Kingdom would display in the "member categories" section, while Category:Former countries in Europe would be displayed as a "subcategory". Unless I'm missing something, this should go a long way towards separating subcategories that just subdivide members of the parent category from subcategories that contain articles about a member of the parent category, to the extent that is a real dichotomy. postdlf (talk) 15:19, 12 December 2010 (UTC)[reply]
That sounds like a good idea. It solves the problem of whether topic categories should be subcategories of set categories and the problem of eponymous category categories. The only problem is there may be examples in which it's not clear-cut. Also, how difficult would this be to achieve? McLerristarr | Mclay1 15:31, 12 December 2010 (UTC)[reply]
Playing games with eponymous articles is an overcomplicated way of solving a relatively simple problem. Topic categories should be explicitly declared as topics, that's all. It doesn't matter whether they're named after an article or not. —Coder Dan (talk) 15:44, 12 December 2010 (UTC)[reply]
That's an interesting starting point. Instead of being on separate pages, sets and topics could be in separate sections of the same page. The sections could be called something like "Content in category "Xs"" (set) and "Content related to category "Xs"" (topic). In the short run though, it's much easier to just agree on a better naming system and add topic or set pages as needed. —Coder Dan (talk) 15:33, 12 December 2010 (UTC)[reply]

categorizing category pages

I was looking at Category:Chaldeans and note that it is in several categories, is this correct? Dougweller (talk) 20:27, 11 December 2010 (UTC)[reply]

Yes. Categories can be in any category they fit in to. However, everything in a category must also fit into its parent categories. For example, Category:The Beatles cannot be in Category:1960s music groups because most of the articles in Category:The Beatles are not 1960s music groups. So Category:Chaldeans is categorised incorrectly but so are a lot of categories. McLerristarr | Mclay1 04:31, 12 December 2010 (UTC)[reply]
I can imagine. Thanks. Dougweller (talk) 06:58, 12 December 2010 (UTC)[reply]
Note that not everyone agrees with the position above that every article in a subcategory must be a member of the parent category as well, as evidenced by the discussion above. There are basically two opinions, the strict "tree" one (cf. McLerristarr above) and the looser "cloud", where subcategory articles are like a cloud of articles related to the parent category, and more loosely to the grandparent category, and so on. Both approaches are being used and being defended and opposed, since both have advantages and disadvantages. The strict one is an utopic situation, you have all the "by nationality" and "by country" categories which can just be thrown away when you take a strict approach (Fasolada is not a "Cuisine by nationality", nor is it a symbol, but if you go up the tree, you will find it in both these categories. Almost every article will be a part of an "incorrect" category when you use the strict approach. Fram (talk) 21:57, 12 December 2010 (UTC)[reply]
I agree. That's why I think we need to have some more consistency in the categorisation system. We need to have better guidelines on the whole topic vs set category thing. We can make exceptions to rule, such as Category:Topics by X, to make categorisation easier. McLerristarr | Mclay1 03:29, 13 December 2010 (UTC)[reply]

What are categories not appropriate for?

There's a discussion at Wikipedia talk:WikiProject Geology#Categorization by discovery date which presents some discussion points that some people here probably would be interested in ascertaining and possibly even comment on. __meco (talk) 10:52, 12 December 2010 (UTC)[reply]

Topic categories and set categories

I discovered a failed proposal for defining category types. It's a bit overcomplicated, but since miscategorisation and the arguments between topic categories and set categories has advanced since 2006, I think it may be possible to revive this proposal, albeit in a very different form. Perhaps we could start tagging some, not every, categories (for pages in the main namespace) with {{Topic category}} and {{Index category}}. For example, Category:Biology is a topic category, whereas Category:People from Yorkshire is an index category (or set category). The templates definitely need work, but I think if we minimise the tagging to clear-cut cases at first, this system should help stop people miscategorising topic categories into set categories. McLerristarr | Mclay1 16:01, 12 December 2010 (UTC)[reply]

"Set" is a better word than "index" I think. It's more widely used and index already has a meaning on Wikipedia. McLerristarr | Mclay1 14:39, 19 December 2010 (UTC)[reply]
I think it is best to leave those templates that were created as part of the failed proposal alone. Instead I have created {{Set category}} and I am applying it to categories. McLerristarr | Mclay1 06:40, 2 January 2011 (UTC)[reply]

By decade categories

I questioned an editor who was moving subcategory category contents in Category:Works by type and year to by decade categories about the need for this. It was suggested that a wider discussion was called for.

The question is do we need to split every by year classification into by decade categories or can we retain the larger group? For many periods, splitting creates subcategories of exactly 10 categories that will not expand or decrease. If there is a need for navigation to surrounding year categories, the decade navigation templates can be used since they do not depend on the category structure, but rather the simple existence of the specific by year categories. There may be a question on how we should we deal with these on the century or millennium basis. Category:Years exists rather well without splitting the years out into smaller groupings. Whatever the outcome in this discussion, it would not prevent by decade or by century categories where the number of items is simply too small to justify creating by year categories for every year. Vegaswikian (talk) 19:37, 14 December 2010 (UTC)[reply]

I'm unable to follow your discussion. __meco (talk) 20:46, 14 December 2010 (UTC)[reply]
  • Comment. I do think that in many cases the decade categories are unnecessary if all they do is hold subcategories for the years in that decade. They are useful when there isn't enough to break a decade down by year, but I'm not sure of their utility if all they do is sit there to complete a scheme. I've also noticed a proliferation of by-millennium categories, which also strikes me as a bit odd. Good Ol’factory (talk) 21:29, 14 December 2010 (UTC)[reply]
  • Often the creation of older works of art may not be assignable with more specificity than a decade or range of years. I think sometimes too art historians use a range to indicate when a painting was painted, not just when it was finished. Otherwise, no, there isn't much point in grouping -by year categories into -by decade categories, given that all of the years in a century can fit in one category display page and will be numerically sorted. Best practice would be to categorize years by decades and by centuries so that way everyone is happy and the decade categories can then accommodate the content that can't be categorized by precise year. postdlf (talk) 22:00, 14 December 2010 (UTC)[reply]

I also have no idea what you're talking about. On a side note, that entire category is WP:Overcategorization. —Coder Dan (talk) 06:00, 15 December 2010 (UTC)[reply]

  • A problem I've noticed with all the year classifications is that the year 2000 is in the 2000s but the 2000s is then classified as being in the 21st century despite the year 2000 actually being part of the 20th century. McLerristarr | Mclay1 06:04, 16 December 2010 (UTC)[reply]
  • there are actually several easily resolved problems with it now because the transition/diffusion was challenged before it was complete, and i didn't see the point of continuing until the major question at hand was settled which is: 1) should all the years be dumped into one; or 2) should they be broken down by decade. imho, the category is too large to for one, and it would be easier to navigate by decade. --emerson7 19:08, 16 December 2010 (UTC)[reply]

Not altogether constructive

I find it symptomatical that the above discussion was initiated subsequent to the below section being created. It shows that our approach towards these chronologies hierarchies is piecemeal and myopic, discussing properties of one single sub-structure that are common to all the similarly structured hierarchies. Nobody has acknowledged the utility of discussing this in a centralized manner, yet a detailed layout for linking between years, decades and centuries is being attempted patched together with no thought being given to how this would impact the larger structure and other subordinate structures from the main years chronology hierarchy. I'm not asserting that the work will detrimentally impact other hierarchies should one model be chosen over another, but going forward in this isolationist fashion is not going to alleviate the impression that mere casual attention is being afforded this work, leaving the seams between the different structures patchy and accidental. And when the next little hierarchy comes along, this process is going to have to be repeated, and some will point to the above discussion and its conclusions as an obvious precedent for how future decisions should be made, and others will point out the different idiosyncrasies in the nature of the subjects being categorized by time, and some people will neglect to consider the finer details being presented and simply vote! to follow the old recipe, possibly leading to nonviable decisions being made that necessitate the entire process having to be redone in the future. __meco (talk) 15:07, 21 December 2010 (UTC)[reply]

Chronologies categories. Proposal for a separate work group

All the many separate and entangled category hierarchies based on chronologies lie within the scope of both WikiProject Years and WikiProject Categories. I was wondering if some people would be interested in a structured effort at streamlining these structures, the way they interact with each other, discussing how year categories should be linked up to decades, centuries and millennia categories, general principles for timeline cutoffs, i.e. when moving back in time and items to categorize become more scattered we should have uniform rules and a guideline for setting (or adjusting) such cutoffs. Also the matter of templates that provide navigating features as well as do much of the categorizing would be of concern. My vision is to have a work group subordinate to both these WikiProjects that would be tasked with addressing and maintaining such chores. I propose discussing this at Wikipedia talk:WikiProject Years. __meco (talk) 21:14, 14 December 2010 (UTC)[reply]

These aren't maintenance categories (and shouldn't be hidden)

Unknown is definitive, missing is temporary. Category:Year of death unknown should therefore not be hidden or considered a maintenance category, as opposed to Category:Year of death missing. This has been discussed before, but I don't know where, so I cannot refer to any previous decisions. __meco (talk) 17:23, 15 December 2010 (UTC)[reply]

I found the opposite opinion at Category talk:Year of death unknown. I can't say I disagree with it (that it's not defining). Looking for the CFD discussion that prompted that. --Kbdank71 17:35, 15 December 2010 (UTC)[reply]
I wrote definitive, not defining. __meco (talk) 17:49, 15 December 2010 (UTC)[reply]
I know what you wrote. I was referring to the opinion at the YoDU talk page. However, speaking to definitive, even if that is so, I don't see how the "unknown" categories should not be considered maintenance categories. --Kbdank71 18:13, 15 December 2010 (UTC)[reply]
How can that possibly be a maintenance category? What possible action could an editor do to remedy this lacking information? The entire distinction between the unknown and missing category lies in whether it is believed that the year of birth can be retrieved from some existing source that has yet to be added to the article's references. __meco (talk) 19:25, 15 December 2010 (UTC)[reply]
What possible action could an editor do to remedy this lacking information? Most likely nothing, so why make it visible? What reader is going to visit Benjamin Agus, for example, and think "You know what, I'm really interested in other people whose date of death is unknown"? Goes back to defining. Nobody is defined by the fact that we don't can't find out when they died. --Kbdank71 20:01, 15 December 2010 (UTC)[reply]
As if Category:1888 births is going to be any more useful to anyone? Should we make all of those hidden also? I believe the rationale for making the missing/unknown categories hidden was their being maintenance categories, not that they wouldn't be of interest to people. And I believe there have been more resounding discussions about these categories than the CFD link below, which frankly provides very little discussion at all. I seem to remember having seen and possibly also participated in discussions on this that were far more substantive than that. __meco (talk) 21:40, 15 December 2010 (UTC)[reply]
Category:1888 births has the potential to be far more useful because people perform research on a particular calendar year quite frequently. I've never heard of research being performed on people whose date of birth is unknown. Good Ol’factory (talk) 21:46, 15 December 2010 (UTC)[reply]
Topics of research don't come into existence based on you or me having heard of them. If we're going to base this on our capacity for conceiving of this or that we'll not be making any good decisions at all. __meco (talk) 22:26, 15 December 2010 (UTC)[reply]
? So you're suggesting that "people whose date of birth is unknown" either has been or could be a valid topic of research? This doesn't necessarily provide the reasons to make decisions, but it does provide a justification for it, because it looks like that's what you're searching for. You suggested there was not a distinction between the two categories, so I just wanted to point out that there almost certainly is. Good Ol’factory (talk) 22:36, 15 December 2010 (UTC)[reply]
the rationale for making the missing/unknown categories hidden was their being maintenance categories Which I agree with, and think they should remain that way. If you can find a discussion that presents the case to change that in a logical way, I may change my mind. --Kbdank71 00:59, 16 December 2010 (UTC)[reply]
Found it: Wikipedia:Categories for discussion/Log/2010 April 14 CFD discussion determined that these categories should go on the article and remain hidden. --Kbdank71 17:40, 15 December 2010 (UTC)[reply]

Well, I have to write this: I knew this will happen. Both sides are right. "Year of birth unknown" is a fact. It's completely different from "Year of birth missing". And yes, it's a valid topic of research..... but research will give nothing! I checked in the past tenths of these articles. Some were mixing "unknown" with "unknown to me" but most of them were about ancient people with no discussion on their death year. I suggested that if we know the century of birth to replace "Year of birth unknown" with "xx-th century births". I don't know what happened with that. "Date of birth unknown" is even worse. Maybe we have to first move on the direction of cleanup. Any volunteers? -- Magioladitis (talk) 01:13, 16 December 2010 (UTC)[reply]

Well, there's the problem, which actually manifested to prompt this entire thread, when the article states for example, "died after 1895". That makes it really hard to add Category:19th-century deaths even. __meco (talk) 11:25, 16 December 2010 (UTC)[reply]
True. Let's start it and we 'll see how many can't be decided. -- Magioladitis (talk) 16:55, 16 December 2010 (UTC)[reply]
Start what? __meco (talk) 18:31, 16 December 2010 (UTC)[reply]
I believe he meant start trying to move articles out of Category:Year of death unknown and into categories like Category:19th-century deaths. For instance, if someone is 70 years old and they were last heard of in 1820, it's safe to put them into Category:19th-century deaths. Then once they are all gone through you can see how many can't be assigned to a century because of problems resulting from the situations such as the example you gave. Good Ol’factory (talk) 01:20, 17 December 2010 (UTC)[reply]
Thanks Good Ol' factory. That's what I meant.
Category:Second millennium deaths anyone? Rich Farmbrough, 15:31, 19 December 2010 (UTC).[reply]

First example I picked has: 

[[:Category:Year of birth unknown]]
[[:Category:Year of death unknown]]
[[:Category:English painters]]
[[:Category:Portrait artists]]
[[:Category:English engravers]]
[[:Category:18th-century births]]
[[:Category:1830s deaths]]

and of course it is still true that "year of death is unknown". Rich Farmbrough, 15:34, 19 December 2010 (UTC).[reply]

We have to give instructions "Give more accurate data as possible". Year is known but not completely unknown. We have some information and we have to use it. -- Magioladitis (talk) 15:42, 19 December 2010 (UTC)[reply]

Brad Evans

Me and Namiba (talk · contribs) have been in discussion here about the inclusion/exclusion of Category:University of California, Irvine alumni from the Brad Evans article. Namiba's rationale is that because Evans is already in Category:UC Irvine Anteaters men's soccer players, he should not be in the parent category; I disagree, and see no harm (actually, I see great benefit) to him being listed in both. Namiba suggested we take the conversation elsewhere for wider input, so any thoughts are more than welcome! Thanks, GiantSnowman 16:50, 27 December 2010 (UTC)[reply]

I agree entirely with Mr. Snowman. Mr. Evans graduated from Irvine and therefore should be included as an alumni. His time as an Anteater has no bearing upon that fact.--EchetusXe 17:58, 27 December 2010 (UTC)[reply]
If no-one else wishes to contribute then I'll be BOLD and restore the category...GiantSnowman 16:51, 2 January 2011 (UTC)[reply]
I'd give it a few more days given the holiday festivities.--TM 16:56, 2 January 2011 (UTC)[reply]
Oh yes, don't get me wrong, I wasn't intending to do it straight away! GiantSnowman 17:02, 2 January 2011 (UTC)[reply]

Question about things going into the categories

Recently I have been noticing that there are some things in various categories that in my opinion ought not go there. For example (and this is just one of many) I noticed today that Category:Dacia articles by quality has the quality assessment documentation embedded into the top of the category. Although I understand the reasoning I think this is probably not appropriate and should have probably been limited to a link to the Project page with that assessment guidance. Should I remove this sort of thing or is this ok to leave there? --Kumioko (talk) 20:45, 27 December 2010 (UTC)[reply]

Generally categories should not have a lot of text. This defeats the purpose of categories by moving the category contents off of the screen. In your example, there is no reason to include the grading criteria there since that is not the place we discuss assessments. They are discussed on the projects assessment page. That's my opinion. Vegaswikian (talk) 20:49, 27 December 2010 (UTC)[reply]

Deletion of PROJCATS cat and populating template

The Category:Redirects from moves and its populating template, {{R from move}} are both up for deletion at Wikipedia:Templates for discussion#Template:R from move. My !vote is to Keep. Is anybody else here interested in keeping this template and cat?
 —  Paine Ellsworth ( CLIMAX )  19:19, 29 December 2010 (UTC)[reply]

Another case to consider

Someone has recently removed skipjack (boat) from Category:Sailboat types and added it to Category:Skipjacks, for which it is obviously the main article. The latter category, however, contains articles on surviving examples of the type (e.g. Minnie V (skipjack)). Meanwhile Category:Skipjacks was moved from Category:Boat types to Category:Sailboat types.

If I'm understanding the system correctly, both of the latter were incorrect, and Category:Skipjacks should be in Category:Boats by type, whereas skipjack (boat) should be returned to Category:Sailboat types and removed from Category:Skipjacks. Is this correct? Mangoe (talk) 21:14, 6 January 2011 (UTC)[reply]

In my opinion, Skipjack (boat) should be in Category:Sailboat types and Category:Skipjacks, and Category:Skipjacks should be in Category:Sailboat types and Category:Boats by type. Although, that's just an observation from quickly looking at the categories. The categories may need seem cleaning up. McLerristarr | Mclay1 02:59, 7 January 2011 (UTC)[reply]

Wikipedia category

{{Wikipedia category}} now has the ability to do everything that {{Hidden category}}, {{Container category}} and {{Tracking category}} can do. Because all hidden categories and tracking categories can be tagged with {{Wikipedia category}}, should we deprecate the use of {{Hidden category}} and {{Tracking category}}? McLerristarr | Mclay1 03:17, 7 January 2011 (UTC)[reply]

Categorization of people

There is currently an interesting discussion going on on Talk:Andrew Wakefield#Correct categorization, please. The question is whether he should be included in (set) categories like the "2011 scandals". As you can see, I'm for the "you have to be one" approach, i.e. since Wakefield isn't a scandal, he doesn't fit into the category. "2011 scandals" is, to be brief, not a people category, it's an event category.

Compare with Category:Political scandals in the United States and the discussion on its talkpage.

Maybe there should be a new policy on categorization of people?

HandsomeFella (talk) 10:16, 9 January 2011 (UTC)[reply]

Hidden categories question

Maybe I've missed something in the talk archives, but I don't understand why hidden categories are visible (not hidden) on pages in the category: namespace. The pages in the category namespace are not necessarily maintenance pages, they are often pages that are intended to convey encyclopedic information to the reader. The reader might explore the encyclopedic information on Wikipedia by browsing the category tree. If he does so, then he won't want to see maintenance categories. So it seems to me that they should be hidden in the category namespace just as they are hidden in the article namespace. Perhaps the reasons for not doing this already are purely technical? Clarification appreciated. 98.118.117.123 (talk) 00:47, 13 January 2011 (UTC)[reply]

Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Categorization&oldid=407572158"