Jump to content

Wikipedia talk:Overcategorization/Archive 1

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 1Archive 2Archive 3→Archive 5

Good idea

This is a very good start. — SMcCandlish [talk] [contrib] ツ 09:21, 24 November 2006 (UTC)

Great work. I support 100%. Kaldari 05:49, 25 November 2006 (UTC)
Yes, it is. Categorization has gotten way out of hand. -- Donald Albury 13:01, 25 November 2006 (UTC)
This is an edit to show support for this proposal becoming a guideline. ·maclean 07:21, 29 November 2006 (UTC)

Excellent - This is a good page to have around. Great work, everyone who's been contributing. -GTBacchus(talk) 08:21, 25 November 2006 (UTC)

Ditto. -- Donald Albury 13:01, 25 November 2006 (UTC)
Great idea. I'm a bit tired of articles that have too many categories. Hopefully this clears many useless categories. Quality over quantity is the key. RobJ1981 19:29, 1 December 2006 (UTC)

Defining characteristic

Jc37 said on the main page about defining characteristics that it "Needs to be clearer, could include any and all categories, depending on one's perspective". I think the last line is a reasonable definition: if you could easily leave something out of a biography, it is not a defining characteristic. I'm bringing this to the talk page to get more feedback on clarifying this. (Radiant) 10:29, 24 November 2006 (UTC)

Sounds good to me so far; that's why I uncommented it. :-) — SMcCandlish [talk] [contrib] ツ 10:38, 24 November 2006 (UTC)
Sounds good to me, as well. -- Donald Albury 13:01, 25 November 2006 (UTC)

Opinion / issue

The point is not only that the person's opinion must be sourced - it must also be relevant to the character. If a newspaper runs an article on how Bill Gates eats a lot of spaghetti, that still doesn't make "spaghetti lovers" a good category for his article. Note that "membership of a political party" is not such a good terminology here since any minor donator can be consired a member even if he doesn't do active work. (Radiant) 10:33, 24 November 2006 (UTC)

I agree on every point. — SMcCandlish [talk] [contrib] ツ 10:39, 24 November 2006 (UTC)
Support. -- Donald Albury 13:01, 25 November 2006 (UTC)

Nominees

I propose reverting the nominees sentence, to discourage all nominee categories. While I think it's just the kind of thing that's suitable in article text, as a category type I think it's unlikely to be of any enyclopedic value, and especially in the case of the lead example (Oscars) it will be unwieldy to the point of collapse. At very best it is only likely to generate woefully incomplete categories that aren't actually useful or informative, just frustrating and misleading. — SMcCandlish [talk] [contrib] ツ 10:49, 24 November 2006 (UTC)

There is not and should not be, a category for Nobel Prize nominees because nominations are secret, and there is no way to verify such nominations. Academy Awards are unusual in that the lists of nominations are a fixed size, winnowed down from some earlier stage of nominations, and very widely publicized and easily verified. Some prizes, such as Grammys, also publicize fixed length lists of nominees, but others don't. Certainly, we should not have lists of nominees for a prize where such lists are open ended and/or not well sourced. I think part is worth more discussion. -- Donald Albury 13:01, 25 November 2006 (UTC)
  • I've changed it to read we should not cat by nominees period, with the oscars as the (to my knowledge) sole exception. Possibly we should look into deleting it if it's as unwieldy as SMC claims. (Radiant) 10:27, 30 November 2006 (UTC)

"Tweak"

Essentially equalled reverting the page.... Any specific reasons? - jc37 11:54, 24 November 2006 (UTC)

  • No, just parts of it. But let me explain - "Tall people" is something that is not objectively defined; that paragraph is not just about the words "famous" and synonyms in cat names, but also "big cities" etc. "Voting independent" is a term people unfamiliar with a two-party system are unlikely to understand. For opinions and awards, see the two sections above. In particular, having a reliable source on someone's opinion does not mean we should categorize by that opinion; being a "member" of a party can mean as little as attending a meeting a year or donating a bit of money; and the suggestion that nominees should never be categorized is contradicted by the apparent fact that they are. (Radiant) 12:19, 24 November 2006 (UTC)

Gender

Circeus asks, What about subcategories of Category:Women? That's a good question :) let's discuss it. (Radiant) 13:01, 24 November 2006 (UTC)

Good start

I'm glad to see this. I've been thinking of doing something similar for quite a while and never got around to it. I'm wondering if this can be made even broader to include most all of the precedents for removing categories. I'd love to see CFD discussion use a list of precedents as the foundation for discussion. Discussion could start with someone saying "X should be removed based on the precedent...." and then the discussion could be about whether the precedent applies, should be an exception in this case, or if the precedent is bad, or policy needs to be changed. There are far to many "votes" based on gut reactions and since few admins are willing to override a majority of "voters" that ignore policy and precedent, many bad categories remain. Those in turn, inspire many more bad categories with the rationale "if xxx can exist, so should yyy". -- Samuel Wantman 23:46, 24 November 2006 (UTC)


One possibilty is to have a different template for CFD discussions. Perhaps something like this:

Category:Anti-Semitic people

  • Problem: This category is a POV magnet. There are many differing views of what an anti-semite is, and it seems impossible to reach a consensus on a NPOV definition.
  • CFD History: Several previous discussions did not reach consensus (links here)
  • Possible solutions:
    • Rename to a NPOV definition Category:People labeled as anti-semites. Precedent: This has worked for categories such as ... (links here)
    • Delete category, create a list with annotations. Precedent: This is the solution disussed in the guidelines for dealing with POV categories. Precedent: Category:Racists...
    • Keep, define criteria for inclusion, and police category. Precedent:This has worked for categories such as... (links here)
  • Proposed changes to guidelines:
    • None -- Precedent applies
    • Allow POV categories, as long as they are labeled as such, and all articles put in the categories have citations to independent sources that document the POV.
  • Extenuating circumstances:
    • This is a constant source of contention between many wikipedians. Previous CFD's have been plagued with vote spamming.
  • Discussion follows:

--- etc...

Anyone would be able to add to this template, but the discussion should be limited to discussiong which precedent applies, or whether policy should be modified and why. Perhaps each discussion should have a moderator/facilitator. There could be a committee of people who are moderators. The moderators should not take part in the discussion, but they could remove comments that are outside of the framework, duplicate other comments, or go off topic. In most cases facilitation is not needed, but in cases where there is not a clear consensus it would probably help. -- Samuel Wantman 23:30, 1 December 2006 (UTC)

The urge to put everything into categories

Human nature, as it shows daily on CfD, is to create non-defining categories again and again. I am afraid it is uphill battle to get rid of them, even with support of an official policy.

I suggest to distinguish categories into primary and secondary, where the primary are strictly regulated as proposed here and the secondary could be used for the less important classification (e.g for the "recipient of ... award" categories). The primary and secondary categories would be visually distinguished (secondary could be hidden by default). It should be possible to add context information like reference to support inclusion of an article in a category.

Example: Josip Broz Tito (the Yugoslavian leader):

Having different types of categories would require support by MediaWiki, though. Pavel Vozenilek 15:25, 25 November 2006 (UTC)

  • Reasonable, but unlikely to happen. A different approach would be to reverse the progress and make a "categories for creation" page (and modify the software so that users cannot simply bypass this, but that's a ten-second tweak). However, that sounds drastic so I'd like to see some statistics first as to how many categories are created each week and how many of those are worth keeping. (Radiant) 12:23, 27 November 2006 (UTC)

Why is this needed?

Why do we need a guideline to say a category is likely to be deleted? This just a fork of Wikipedia:Categorisation. Tim! 15:41, 25 November 2006 (UTC)

The same as the value of any essay or guideline. It's something someone can point to and say "this is what I mean", so that it doesn't have to be written out every time when commenting in CfD, for one thing. - jc37 15:44, 25 November 2006 (UTC)
Why do we need this specific guideline though? Wikipedia:Categorisation should be a single comprehensive guideline to categorisation. Fragmenting it obfuscates any benefit of a guideline. Tim! 15:47, 25 November 2006 (UTC)
I think this compares to WP:PEREN, only as Perennial (re-)creations. - jc37 15:52, 25 November 2006 (UTC)
WP:CAT doesn't address the same ideas as this proposal does. It is also long and sprawling, resembles a scribble pad with all those lines struck out. It needs shaping up and a partial re-write if it is to make this page redundant. Unless and until that happens, this proposal will be useful to have as a guideline. This proposal reflects current practice, clearly explains common category downfalls, and does it succinctly. While possibly too detailed and not worded towards being a policy, this would make an excellent guideline in my opinion and (possibly) a good addition to the manual of style. And yes, it is sort of analagous to WP:PEREN for categories instead of policy proposals. BigNate37(T) 19:21, 25 November 2006 (UTC)

Triple intersection

I don't think triple intersection per se is the problem, more intersection of unrelated topics. Obivously the more intersections you do the more likely you are to result in a silly category, but there are quite a few examples of useful triples: American film actors is fine, because film and acting are very closely related. Tim! 17:55, 28 November 2006 (UTC)

This popped out at me as well. The problem with intersections is not so much that they exist, but that the parents often get diffused into them. Instead of having meaningful topic indexes we end up with oodles of subcategories that are more esoteric and obscure. This weakens the usefulness of the entire categorization structure. So the important criteria is usefulness. This would be less of an issue if multiple intersections did not diffuse their parents. The other problem with multiple intersections is that they can add a very large number of categories to an article. So as long as there is a possibility of making an exception to the precedent of triple intersection in cases that clearly have utility, we should do everything we can to discourage them. -- Samuel Wantman 18:33, 28 November 2006 (UTC)
And myself as well. I can think of several cases where triple intersection is not only viable, but badly, badly needed. I realize many of these cases fall into "narrowing of a trait", but that gets fuzzy very quickly. Instead of setting the bar there, I'd like to see it changed to something similar to "arbitrary intersection". Crystallina 19:48, 29 November 2006 (UTC)
Reworded, please tell me what you think. I can't think of any useful triple intersections myself (off the top of my head) but there are several things that may look like triples but in fact aren't (I'd consider "film actor" to be a single profession, not two). "arbitrary intersection" as Crystallina suggests is also something to avoid. (Radiant) 10:36, 30 November 2006 (UTC)
I would like to hear more opinions on whether "Italian-American <foo>" is all that useful a categorization. Maybe it is in fact a triple intersect (1.foo, 2.in America, 3.of Italian origin). (Radiant) 16:25, 30 November 2006 (UTC)
I consider film actors to be an intersection of actors and film or film people, but it is a bit fuzzy, maybe a 2.5 intersection :) Tim! 17:49, 30 November 2006 (UTC)
  • Perhaps we should say that triple intersections are usually not a good idea, but in the proper setup they can be useful. After all, nobody's proposing to speedy the lot of them. (Radiant) 08:47, 1 December 2006 (UTC)
I'm very skeptical of this proposed principle (and my inner pedant wants to point out that these are really "double intersections", given that an intersection necessarily involves two sets). Guidelines are supposed to codify existing practice, and I see no groundswell against Category:US Democratic Party politicians (obvious NC issue aside) or Category:American jazz pianists. The logic of particular intersections should obviously be justified, though. Degree of intersection and whether or not to "diffuse" also seem to me to be entirely separate questions, but while categories get ever-larger, and while the software doesn't support intersection, diffusion seems to me to be a practical necessity. (On the latter, while arbitrary user-queried intersection would probably be the straw the melted the server backbone, some sort of software support would be both feasible and desirable, though it would probably not not entirely eliminated the need for "manual intersection", unless you're going to have union categories in software, too.) Alai 23:53, 1 December 2006 (UTC)

Good overall. Specific comments

Looks good overall. Some specific comments -

  • Non-defining or trivial characteristics - My only concern here is that "non-defining" is somewhat vague. Certainly only facts verified in the article itself should be considered when categorizing, so if a fact isn't even worth mentioning in the person's article it likewise shouldn't be used to categorize them. But on occasion I'll see a category that some claim represents an important, defining trait while others don't. As an example, Category:Fictional mutes came up recently. I would consider muteness to be a pretty significant, defining characteristic for a fictional character, putting me in disagreement with Radiant in that cfd. The point, though, is that the phrase "non-defining" might be overly vague, and perhaps there is a more specific, objective set of criteria that could be used to narrow down the concept.
  • Award winners and nominees - The proposed guideline says that the rule of thumb is if an award doesn't have an article, then a person should not be categorized by receiving it. I would go a step further, and say that most awards should probably use lists instead of categories. There are many, many awards in the world, all of which consider themselves notable and which contain a great deal of overlap and redundancy. From a search standpoint, a user can just as easily find a specific award winner from a list article as from a category page, and likewise can tell by looking at someone's article if they won a specific award. So for the most part, categorizing by awards isn't actually needed, and might possibly create a problem with redundant or less significant awards creating excessive categories on a biographical article. Therefore I'd suggest considering limiting award categories to only those awards which are particularly prestigious and represent pinnacle lifetime achievements for the winners. (In other words, use similar guidelines for awards as for "Inclusion in a published list")
    I strongly concur that the categories for "winner of X award" just add unnecessary clutter. Especially with, for instance, prolific actors, prestigious scientists, and so on, they will have lots of awards. Lists are much better -- the person can have the list of their awards, and the awards can have the list of their award-winners. --lquilter 14:40, 10 January 2007 (UTC)
  • Actors by film, or films by actor - I agree with this guideline, and in fact I feel the same reasoning also applies to television shows and video games. Most television programs and video games do not need their own unique categories for actor lists, and in fact list articles are a better way to present cast lists for those shows. I would suggest expanding this guideline to those media types as well. Dugwiki 18:58, 28 November 2006 (UTC)

(resulting subdiscussion moved to a seperate thread) Dugwiki 17:14, 1 December 2006 (UTC)

  • "Non-defining" needs clarification. Note that for whatever definition we give here, there will always be borderline cases that need to be discussed. I would say that "mute" is such a borderline case. I've narrowed down the 'awards' section per your suggestion. I tend to agree with what you say about actor lists for cvg, but note that most TV shows do in fact have a category for "cast and crew".


It seems clear that the topic of how to handle "actor-by-series" categories is pretty complicated in its own right. To that end, I'll start a new thread on this talk page specifically for that topic. I think having a seperate thread specifically for those issues will help keep those discussions in one place that's easy to find and review. Dugwiki 17:11, 1 December 2006 (UTC)

Occupations by city

What are peoples thoughts on occupation by cities categories? I'm thinking of issues such as Male models from Dallas, or the more recent Category:Artists from Pittsburgh. It seems to me that it's inappropriate to create occupation subcats at the city level. I have no particular objection to location / occupation cats, but only when a national category is getting very large. For example, by the time I created Category:American musical groups by state there were already about 10 state cats and hundreds of members in the national cat. -- ProveIt (talk) 02:35, 29 November 2006 (UTC)

I don't have a problem with divisions by city, provided that the category is defined in such a way that the article will only have one or two such listings. For example, if you define "Artists from Pittsburgh" as "artists who were born in Pittsburgh or who notably lived in Pittsburgh", then you wouldn't be likely to have more than one or two city-specific artist listings. On the other hand, if you defined it as "artists who have toured in Pittsburgh", then you could end up with the artist having a hundred such categories, one per city they tour.
So basically as long as the city specific categories are defined in such a way as to limit how many of them apply to an article, then it's ok. And in cases where the state-specific divisions are particularly large, subdividing by city or county might prove useful. Dugwiki 17:05, 29 November 2006 (UTC)
I have always found these occupation by cities categories to be a misguided attempt find a commonality between people. What does being an artist have to do with being from Pittsburgh? It implies that there is some Pittsburgh style of art, or Dallas style of male modelling. I can understand categorizing people by where they were "from" but why subcategorize those people by occupation? Those listed in 'Cat:1975 Births in Pittsburgh' would have a more common background (also it can't be duplicated: you're born once in one place, though you can have many occupations). ·maclean 19:14, 29 November 2006 (UTC)
Locality can, I think, influence artists. For example, New Orleans is known for its jazz, and Greenwich Village for its bohemian culture. In general, I think it's reasonable to believe that artists in the same local community can share local preferences and tastes and will produce works with some related similiarties. From a reader standpoint, I can also see how visitors to Wikipedia looking to read about a particular city or region would be interested to read the biographies of that city's cultural figures. Dugwiki 22:17, 29 November 2006 (UTC)
The "influence" arguement keeps appearing in the CfDs and is the reason why I call the categorization misguided. If the categories were about common influences then I would be all for it, but that is not what these categories are saying. They merely (and vaguely) imply that being from the city makes them similar as artists (or other occupation) without saying how or why. So what is the influence of Pittsburgh on writers? or the Pittsburgh school of art?
Perhaps this is similar to when all the Cats:culture-by-country were re-named to Cats:culture-by-nationality (it is not Category:Artists from Norway, it is Category:Norwegian artists). Likewise for occupations Category:Norwegian lawyers, not Category:Lawyers from Norway.
Beyond the "influence" arguement it is a trivial connection of people being "from" the same place. Yeah, this might be interesting from the reader's standpoint. But I think it is misleading for an encyclopedia to officially categorize them like that. ·maclean 03:54, 30 November 2006 (UTC)
  • I think we should stop at the state level here and not characterize people by "city and profession" (with a possible exception if the city does indeed have a distinct style). We may have to drop some examples on CFD to see if there's precedent for this, though. (Radiant) 10:44, 30 November 2006 (UTC)

Turning this into a guideline

If each section were to be linked to a few CfD discussions which resulted in deletions for the reasons stated in each section, then this would not be the creation of anything new, but a statement of current practice. As such, it doesn't need to go through any community approval process, because the page would simply be a guideline based on past precedent. I think this can easily be done. I'd do it myself, but I'm leaving town for 10 days in a few hours. -- Samuel Wantman 07:51, 29 November 2006 (UTC)

For those who can read French (and are quite masochist)

We are presently having an awfully chaotic discussion on :fr about the sole question of overcategorization linked to nationality or ethnic origin. Not so broad as your debate, but we lost much more energy. Still running at fr:Wikipédia:Prise de décision/Catégories par nationalité, identité régionale, ethnie... French Tourist 23:07, 29 November 2006 (UTC)

Well, don't be so rude! The discussion has flown from misunderstanding about what is relevant in fr.wikipedia in term of national categorisation. Hard to find a consensus! There is a major word: relevance, but without a conclusive definition. So let's see if we can decide. Gwalarn 12:50, 1 December 2006 (UTC)

The tree organisation and a key word: relevance

Everybody hopes to find relevant categories in an encyclopedia. But relevance relates to personal opinion: I may find relevant a category for historical books written in the XIXth century about French regionalism mouvement, or some use of technology in Middle-Ages; it is perfectly encyclopaedic, but I guess my interests are not common.

The only way to create a category is to put it as a node in a hierarchal tree. So all categories are "top-down". If I look after "American writers" in the XXth century, I must create "American writer in XXth century". If I want them by origin (Native, African, White, etc.) or by style (poetry, politics, science-fiction, etc.), it is quite impossible to find (otherwise create this: "Political American writer in XXth century from African origin", 61 characters).

An other way will be to give access to a cross-category finder. I don't know if it exists on en.wikipedia, but it exists on the French one. Look at: CatCroiseur. It is all new, but for me this is the major improvement in the way of how to (well) categorize. Gwalarn 13:13, 1 December 2006 (UTC)

Note: The category system is not a "tree-based" system (though it can be used as one). See WP:CAT#Categories do not form a tree, for more information. - jc37 11:04, 30 December 2006 (UTC)

The "not a tree" is referring to how no cycles are present, and that multiple starting schemes are allowed to interact. The "direct" part in "Directed acyclic graph" does implies that articles should be in the most precise category, if I understand the article correctly. Of course, this should be exercized with care, as categories are of 2 completely different scope types (topical, and taxonomical), and there are at least 3 significantly different subcategorization schemes (full division, subset, geographical).
This quite often leads to a need for sorting a category or article in two categories, one of which is is an ancestor to the other.Circeus 16:44, 30 December 2006 (UTC)

Something that may help

Suppose we had the following...

  • Categories can be "approved" or "unapproved" (we may need a better name for that).
  • At the bottom of an article, unapproved categories have a different markup (so the monobook.css can be employed to e.g. give them a different color).
  • There's no other effect to being "approved".
  • Categories start as unapproved; admins can flag them approved, or unflag them.

Many categories already exist according to certain schemas or hierarchies, e.g. the "<country> <profession>" cats, or many cats by year. We can quickly blanket-approve them. This way, new schemas and discrepancies from schemas are clear, and this makes it easier to standardize issues and locate problems. Yes, this requires a software change, but it would be an easy one. (Radiant) 14:10, 1 December 2006 (UTC)

I agree that "approved" might not be the right word. Also, there have been several proposals that I've read, talking about "minor" categories, and other similar ideas.
If your suggestion were implemented, I presume it would be something like:
  • Modify code so that all new (non-approved by default) categories show up on a page as a different colour to admins, and don't appear at all to non-admins.
  • Create WP:RFCA (Request for category addition, or some other such name), which would be a CfD-

like system, which would be discussions about adding a new category. Categories for speedy renaming (and possibly all category renaming nominations) would probably be merged with this process.

  • Once approved (per closed discussion), the category could be "tagged" by an admin in some way so that it would then appear normally on pages. Programming-wise, this would be just be setting a flag on the category, which I presume "should" be relatively simple, and giving the admins the ability to "set" this flag.
I honestly don't have a problem with the above. I think it would help quite a bit. Anyone can still create a category, but since there would be a step to go through before to actually seeing the category appear, there might be less of a wish by vandals, and others to create "less-than-useful" categories. - jc37 14:38, 1 December 2006 (UTC)
Considering that it works exactly the way CfD works, and different coloured links is exactly how admins see "other material", I just don't see this as "against the way wikipedia works". If anything, it embraces how other processes currently "work". - jc37 17:45, 1 December 2006 (UTC)
No, CFD processes the minority of incorrect categories, this would require processing the majority of the thousands of valid categories. Tim! 17:56, 1 December 2006 (UTC)
Which means that your concern is about a numeric quantity, and not any of what you just said above. And of the "2000" mentioned above, how many do you think will be "speedied" along, and not requiring further discussion? This simply is a positive, proactive way to deal with this. - jc37 18:09, 1 December 2006 (UTC)
No it is a poorly thought through overly-bureatic solution to a problem which does not exist. Tim! 18:38, 1 December 2006 (UTC)
"...poorly thought through..." - rather than respond yet, I think I'll give you the opportunity to strike out (for rather obvious reasons) at least that part of your comments : ) - jc37 19:57, 1 December 2006 (UTC)
No it is poorly thought through, the backlog of aprroving several thousand categories a week doesn't bear thinking about. It is also against the spirit of other guidelines such as WP:BB and is an attempt to reverse deletion policy by making there be a requirement for having consensus to have something, rather than consensus to delete. Tim! 20:16, 1 December 2006 (UTC)
Well, let's rattle off a few acronyms... WP:AGF, WP:CIVIL, etc... I understand that you don't like the idea, and I understand that you have a perspective. But I just might suggest that you consider your tone. As for the "reversal", then by such a standard, RfA shouldn't exist, but rather everyone should be an admin, and there should only be a system to "de-sysop". - jc37 20:23, 1 December 2006 (UTC)
The RFA comparison is not valid because it is to evaluate who is competent to wield potentially destructive tools such as delete, protect and block. A more valid comparison would be with the ability to create a user account, which would only be allowed if a user promised to behave and then a consensus gathering exercise would be carried out to see if that user be allowed to edit. Tim! 20:33, 1 December 2006 (UTC)

I imagine there would be large resistance to this (just look at the grief WP:WSS/P gets, for what's in theory just a process internal to a wikiproject). But what about a "kinder and gentler" version of this, along the lines of category-page, or category-talk-page markup citing cat naming convention, CFD, or other "precedent" for why this category exists, is named as it is, etc? Just for the purposes of tracking the provenance of a given category, and cutting down on double-takes. Alai 01:05, 2 December 2006 (UTC)

Actor-by-series and actor-by-film categories

I have moved the previous discussions to date for the "actor-by-series" categories here. Because this topic is pretty complicated and maybe a little divisive (editors have expressed strong opinions on both sides), I think it's a good idea to have it take place in its own thread. That way anyone interested in contributing to the discussion can find it more easily. Dugwiki 17:17, 1 December 2006 (UTC)

  • My point on the cast and crew lists, though, is that most TV shows don't actually need a category for that. Listing the cast and crew in the article or in a seperate castlist article accomplishes the same goal, and avoids the problem of having every single show an actor has appeared in appear as a category in their article. Thus, in my opinion, most of the cast and crew categories for specific shows should ultimately be deleted. Dugwiki 16:47, 30 November 2006 (UTC)
  • Or to put it another way, I don't see an important distinction here between creating a unique category for actors in a specific film (eg "Frankenstein Unbound actors") versus creating a unique category for actors in a specific tv series (eg "Black Scorpion (TV series) actors") (using Roger Corman for my hypothetical examples). Both films and actors typically have large cast lists, and many actors appear in numerous films and numerous television series. But the proposed guidelines signal out specifically films while leaving television series alone. Maybe I'm overlooking something, but why would they be treated differently when it comes to category creation? Dugwiki 16:55, 30 November 2006 (UTC)
  • I understand that they are currently being handled differently because of a lack of consensus on the TV side. What I'm not clear on is why, logically, they should be treated differently in Wiki category guidelines. If actors-by-TV series categories are considered generally ok, for example, then why wouldn't you by extension also consider actor-by-film categories acceptable? Or vice versa, if the guideline is that actor-by-film categories are a bad idea, why wouldn't the same reasoning apply to actor-by-tv series? Lack of consensus on one side doesn't explain the reasoning; it just indicates there is an inconsistency that probably should be examined. Dugwiki 18:15, 30 November 2006 (UTC)
  • Even assuming you're correct about television series being somehow "more defining" than movie roles for larger number of actors, I would think that only apply to TV series that are somehow "defining" of a career. And it still leaves a somewhat inconsistent policy. For example, Tom Hanks got his start on Bosom Buddies, and later had a major break out role in Big and A League of Their Own. But even though (in my opinion) those two movie roles are more historically important, the way the proposed guidelines read it would be acceptable to make "Category:Bosom Buddies actors" but not acceptable to make "Category:Big actors" or "Category:A League of Their Own actors".
  • So if the criteria here is whether or not appearing in a tv series is "defining", I think that should be spelled out. Otherwise basically any television show that has an article could argue in favor of having its own unique category, while even major, important films are left with a very high burden to have the same categorization scheme. Dugwiki 20:41, 30 November 2006 (UTC)
  • Also, as we've discussed briefly elsewhere Tim! and with others, there are some technical problems to deal with in TV actor categories that probably should be considered in a general form under this policy. One issue is voice acting. Voice actors in particular are quite prolific; Tara Strong for example has over 200 film and TV credits on her IMDB listing. A high bar might need to be set before allowing animated series to have their own categories, lest you end up with too many such categories on the related voice actor's articles. Similarly, there is the question of whether or not to include guest appearances in television categories. It seems like the consensus is leaning against guest spots, but at the same time leaving out guests might make a category of actors incomplete, not to mention there being some categories that specify "regulars only" and others that don't.
  • So regardless of which way the discussion goes, I'd suggest that this proposed Overcategorization policy might be a good place to try and reach some consensus on these TV actor issues and spell it out for consistency. Dugwiki 20:50, 30 November 2006 (UTC)
  • I think the "performers by performance" categories should probably all go. However, currently (as noted above), TV series is being treated as an exception to that. I don't see why this can't be specified, and just saying that TV series cast members is currently an exception (compare to the academy award nominees). I also think that if we allow TV series to be an exception, we should allow Radio series cast members to also be an exception, else it's a double standard. (Fibber McGee and Molly, or the Jack Benny show, for just a few examples). - jc37 15:22, 1 December 2006 (UTC)

Let me try and lay out some of the basic questions here that should probably be addressed by any policy dealing with actor-by-series categories.

  • Should actors-by-film and actors-by-tv-series categories be handled differently? If so, why? They are, I think, clearly very similar in nature, but some people feel that TV series are inherently different than films in how they "define" an actor's career thus should be treated differently.
  • Which television series should be allowed to have their own actor-by-tv-series categories? All of them (including things like Bosom Buddies? None? Only shows that are deemed "really, really notable"? Only shows that have multiple spin-offs and movie tie-ins (eg Star Trek actors)? Should animated tv series be treated differently than live action, because of the prolific nature of voice acting? What limits, if any, should be placed on which actor-by-tv-series categories are acceptable?
  • What about other media, such as radio series or stage? Should actors-by-radio-series or actors-by-original-stage-production be allowed?

These seem to be some of the fundamental questions that the policy should address, if possible. Feel free to add thoughts, or add additional questions if I overlooked something. Dugwiki 17:29, 1 December 2006 (UTC)

Actors by stage is different, because each production is (roughly) the same. (Though comparable to film.) - jc37 17:45, 1 December 2006 (UTC)

Yeah, I don't think actors-by-stage is a good idea, but I've seen a few editors argue in favor of it periodically. That's the main reason I mentioned it. Dugwiki 17:59, 1 December 2006 (UTC)

I can only say that in general when browsing trough the TV articles, I have found these categories to be very useful. Much more useful then the "2000s american television series" categories for instance. Also the primary cast of actors in a television series is usually much more elemental to the television series then to a film. A film is usually carried by 2-4 actors. A television series is carried by 4 to (especially as of late) 16 characters/actors. A film usually has importance for 2 years, whereas a series might have importance for over 10 years. All that time shaping it's characters and trough that also it actors. Note that whereas film actors like to say "it was almost a family" quite a lot in their interviews, TV series usually tend to BE families of actors (You should see the final special on "That 70's show"). Now i'm opposed to having every series have such a category however. It's like with having pages for the individual episodes of a television series. The series needs to reach some level of popularity and notability before such a thing becomes useful. TheDJ (talk • contribs • WikiProject Television) 20:45, 20 December 2006 (UTC)

  • The discussions here keep going to things like "there needs to be super-notability" and not every TV show should have such a category. That requires a type of judgment that can't possibly be reflected in the category name itself, and can't possibly be easily maintained. --lquilter 14:46, 10 January 2007 (UTC)
  • Here's a little perspective. I've identified at least 242 series in the subcategories of the Actors By Television Series category (there are a few more in the form of subsubsubcategories). From a random sampling (I deliberately excluded Law & Order for reasons mentioned below) of 23 series, I found an average of 131 actors in each category. This accounts for very close to 32,000 categorizations. Have television series actually produced 32,000 career defining roles of notable actors!?
This excluded consideration of Law & Order, and its spinoffs, because I didn't want to distort the figures for the sake of this one enormous group of categories. The 5 Law and Order Series, according to the current form of Wikipedia, have produced 1551 career defining roles of notable actors.
Soon Wikipedia can change its name to "Encyclopedia of Television Actors."
Some members may favor renaming these categories to reflect "cast" lists rather than "actors" lists, and hope to enforce nebulous concepts regarding who is part of the primary cast of a series, and which roles are career defining... I hope you're prepared for the upcoming 32,000 debates on this subject. Assuming that this categorization will really stop at the current list of 242 series.
But I say delete them.zadignose 04:56, 15 January 2007 (UTC)
Very good point with experiential data! I think this strongly supports the point that the categories, whether called "actors", "cast", or what-have-you, cannot be maintained as simply "career-defining roles". Delete them all! And convert to lists. --lquilter 23:01, 18 January 2007 (UTC)
  • I'm a newbie, so please excuse the ignorance behind this question, but why can't these "categories" be dealt with as lists? From what I've seen, lists are under-utilized (at least when it comes to TV-related entries), which results in overcategorization. (I am still trying to understand why individual TV episodes merit their own articles, but that's a whole other debate.) And trying to determine a "notability" standard is far too subjective.--Vbd 08:23, 15 January 2007 (UTC)

Jumping in late: First, I think "cast" is much better than "actors". It should be restricted to those with a regular or important recurring characters, not one-time appearances. There are rare exceptions to this: For Star Trek, even one-time appearances change an actor's career for life.

Second, I think a show has to have some enough cultural significance to affect the careers of the actors on it. A show that is only on for one or two seasons, even if it's excellent, is just a stepping stone to the next job. For example, Dark Angel was an excellent show, but all of the principals have gone on to better things. OTOH, a show like The West Wing puts a really indelible mark on its primary actors, even people like Rob Lowe who would much prefer not to be defined by it.

If we can pull a consensus out of this discussion (which I'm sort of hearing), I think it will be more broadly accepted. Replacing categories with lists will work for most people. Avt tor 22:32, 18 January 2007 (UTC)

Possible merge to avoid instruction creep

Just learned about this proposal today, and I really like a lot of what it says. The only thing is I worry a lot about the raw number of guidelines we have on things. The most obvious thing to do would be to merge with Wikipedia:Categorization, but that's a large guideline in itself. I suggest that we do merge the guideline and proposed guideline (once accepted, that is), but also simplify a lot of stuff in Wikipedia:Categorization or at least move the non-vital things to a second page. The points made on this proposal are so good that I think they shouldn't be on a second page, but rather on the first. Thoughts? -- Ned Scott 22:05, 1 December 2006 (UTC)

Arbitrary geographical boundaries

We seem to have a lot of bad categories that use arbitrary geographical boundaries to classify things, partcularly people. Pastorwayne's multitude of clergy-related categories (such as "Bishops from Ohio") are good examples but not the only examples (for example, note "Quarterbacks from Louisiana"). While the substance of this categorization is already covered in WP:OVERCAT, this specific type of overcategorization deserves to be noted. Dr. Submillimeter 12:56, 15 December 2006 (UTC)

Arbitrary geographical inclusion of people
Examples: Roman Catholic Bishops from Ohio, Quarterbacks from Louisiana, Male models from Dallas, Texas
This is a variant of both arbitrary intersection and arbitrary inclusion limit where an arbitrary geographical boundary is applied to specific groups of people even though the boundary does not have any major bearing on the individuals' careers or accomplishments. Taken to an extreme, such categories may also be interpreted as including anyone who had ever lived or worked within the geographical boundary, which limits the categories' effectiveness in bringing together similar articles. However, geographical boundaries are useful for dividing people into regions that are directly related to their careers or accomplishments (for example, Roman Catholic Bishops of the Diocese of Columbus, Ohio or the New Orleans Saints quarterbacks).

A few additions

"non-defining characteristic" should not only apply to people. Added "sexual preference" to race/religion section. Changed "triple intersection" to "overly narrow", and reworded. Added two specific examples. Please copyed. (Radiant) 13:14, 15 December 2006 (UTC)

Consensus?

I'm not sure there's a consensus for this as a guideline yet, we probably need more discussion on it. In particular, an issue with the situation was brought up on the mailing list earlier this week about this exact issue regarding placng articles in subcats as well as the parent cat, and there seemed to be some pretty robust discussion on the issue, telling me that there's really no agreement at this point on overcategorization. As there aren't binding precedents here, I'm not seeing consensus here, so thus the reversion. --badlydrawnjeff talk 14:22, 18 December 2006 (UTC)

  • Yes, there is, as evidenced from the reactions to this page, the fact that it's being cited quite a lot on CFD, and the fact that it's based on actual practice. Please give a link to that mailing list discussion? (Radiant) 14:39, 18 December 2006 (UTC)
    • I don't see it on the page, citing at CfD is irrelevant, and if actual practice is wrong, actual practice must change. Here's the relevant thread. --badlydrawnjeff talk 14:42, 18 December 2006 (UTC)
      • That thread is about a novice user who is dissatisfied with the guideline that an article should only be in the most specific category, and several other users explaining why we have that guideline. That has no bearing on the content of this page, which is about categories we shouldn't be having in the first place. If you don't like actual practice you are free to propose a change to it, but dislike of practice is not a valid argument against writing down that practice for easier reference. (Radiant) 14:49, 18 December 2006 (UTC)

Right now, Badlydrawnjeff, you are the sole user that has been arguing this guideline is not based on consensus. Your arguments have nothing to do with either the building process or theexistence of consensus. If you can't disagree with something in the guideline, you are really trying to force an open door. Circeus 18:21, 18 December 2006 (UTC)

Does Wikipedia have a process by which we could vote on this becoming a guideline? (I would support this page as a guideline.) Dr. Submillimeter 18:25, 18 December 2006 (UTC)
Whether I'm forcing an open door or not is irrelevant (especially since this hasn't been advertised anywhere that I'm aware of), but, since you asked, I don't agree with the essay at all and don't care to see it as anything actionable or binding. The essay is based on the opinion that a wealth of categories is unhelpful, which I strongly disagree with. I take the opposite tack - the fewer the categories, the less helpful they can be due to people looking for specific information. --badlydrawnjeff talk 18:39, 18 December 2006 (UTC)

I'm not following this thread at all. I read the e-mail post and agree with everything it said, but it sounds to me that the e-mail writer is complaining about overcategorization and then the diffusion of larger categories into more and more narrowly designed categories. The post says "Now, as the categories become more and more diluted, this [useful populated categories] is becoming more and more unavailable." This is exactly the reason why I support this page. Am I missing something here? -- Samuel Wantman 02:47, 19 December 2006 (UTC)

  • Jeff, since you're the sole dissenter here (and that mailing list thread does not support your statement, but opposes it), I would like you to opint out anything on this page that is not in fact common practice. Guidelines are descriptive, not prescriptive; so since this is an accurate description of common practice, it is a guideline. Your sole dissent does not indicate a dispute or controversy. (Radiant) 12:14, 21 December 2006 (UTC)
    • I don't know if any of it is truly common practice. You claim that guidelines are descriptive, this continually fails to be true. If you want this to be a guideline, demonstrate the wide consensus, advertise it a bit, and get discussion. The mailing list shows some serious discussion about how categorization should work, and to claim that it opposes my position, I believe, is false. If it's such an indicator of common practrice, you'll have no problem advertising it at the pumps and having people come here and approve of your position on the matter. --badlydrawnjeff talk 13:43, 21 December 2006 (UTC)
  • Just incase it needs to be said again, I think this is a great idea. I'd say something more, but everyone else has pretty much summed it up. (if you really want me to, I'll be glad to go into detail why I support this) -- Ned Scott 12:26, 21 December 2006 (UTC)
    I too think that this is a great idea (we've been discussing such things on CfD, without a centralised page, for some time). However, I think that the page still needs some work to truely represent consensus, before it's labelled a guideline. - jc37 12:56, 21 December 2006 (UTC)

Jeff, you have so far failed to provide either support against the policy (If so many people disagree, why are you the only one clamoring here?) or any relevant counterexamples. If you can't either argue on the principles or the practices, then stop with the revert war. We want arguments, not rhetorics.

  • You claim that guidelines are descriptive, this continually fails to be true
    • How so? How about providing some, you know, examples?
  • The mailing list shows some serious discussion about how categorization should work, and to claim that it opposes my position, I believe, is false.
    • Nobody claimed it "opposes" your view. Most, however, noticed that it was completely unrelated.
  • you'll have no problem advertising it at the pumps and having people come here and approve of your position on the matter.

Numerous people have edited the guideline to help clarify it. You have not even TRIED to objectively explain what was wrong with it, and tried to pass it off as a lapse in process, forgetting that Wikipedia is not a bureaucraty and will most likely NEVER have a formal process beyond mere codification of consensus to establish them.Circeus 21:51, 21 December 2006 (UTC)

To answer your questions:
If guidelines were descriptive, we'd allow for more fair use and include more articles at AfD (given the spate of reality show AfDs).
No, it was clearly said above that the "mailing list thread does not support your statement, but opposes it." I disagree with that and the idea that the two are unrelated.
Okay, so it was at the policy pump and a couple irrelevant archives. A great start.
What's wrong with it is simple - it makes our categorization system less useful by suggesting to lessen the categories, when it's a great way to expand the information. This has nothing to do with a bureaucracy and everything to do with actually finding consensus instead of asserting it's there, as the tendency by certain editors continues to be. --badlydrawnjeff talk 00:28, 22 December 2006 (UTC)
So let me understand what you are saying. You are objecting that many categories such as the examples mentioned both real and hypothetical should not be deleted. This seems to be a different issue than whether these categories do routinely get deleted at CFD, and usually for the the reasons that have been written in this guideline. If you are saying that these categories do not routinely get deleted we can easily determine the truth of the matter by looking at the CFD logs. If you are saying that the common practice is an error, we can add language that explains why people such as you object to removing categories for these reasons. I have no problem with either way of going.
BTW, when I first started spending time at CFD I would bristle every time I read Radiant!'s comments that said "Delete - overcategorization". What looked like his sometimes misguided zeal at removing categories was one of the reasons I started participating. Sometimes I still disagree with him, but as more and more categories get created and large ones get diffused into little ones, I find that I have become the person writing "Delete - overcategorization".
There is a basic decision that we all have to make about categorization. It is whether we see categorization as a system of tagging in which a tag may be trivial or important, or as multiple taxonomies in which we are trying to construct an intelligent map of knowledge. As this is a project to create an encyclopedia I think we are doing the second. If you agree, that means that we have to draw a line somewhere between things we think are important and those we think are not. With few limitations, you can add information to articles and lists. I don't think we want to add categories for every possible way of classifying an article, at least not with the software we have now. Perhaps the software will evolve so that each article gets hidden database entries that allow for all sorts of database analysis. Until that day, I think we have to take a practical approach to what our categorization system can and cannot do. -- Samuel Wantman 04:52, 22 December 2006 (UTC)
See, I'm with you on the second. The problem is that I believe that, to achieve the second, a misguided essay like this gets in the way. --badlydrawnjeff talk 11:40, 22 December 2006 (UTC)
  • Except, of course, this is neither misguided nor an essay, and describing a page people are cooperating on in negative terms doesn't really help your case. I would be happy to hear your constructive thoughts on making an intelligent map of knowledge. >Radiant< 12:34, 22 December 2006 (UTC)
Let's build on what we agree on. We all seem to agree that the categorization system is an attempt to make an intelligent map of knowledge. The problem that this page tries to address, is getting people up to speed on the community decisions about that map. I've been at CFD for about two years, and at times it seems to be a huge energy sink. Many discussions happen over and over, and categories get kept or discarded without talking about or deciding the underlying criteria for what makes a good useful category. I recycle some of my comments over and over about why we have too many eponymous categories, or have chopped up good categories into trivial intersections. I want a page, like this one where I can simply link with a short comment that I think it is an overcategorization, and the past history of CFD shows that the community has come to the same conclusion. The only problem with the page that I now see is that it is lacking the citations that would make it verifiable. There should be links to the CFD cases that show the consensus to delete categories like this. Not having a page like this is like having a judge who must write new laws every time he hears a case. I see this page as being helpful to educate users about categorization, how CFD works, and the accumulated wisdom of the community. If people disagree with any of these overcategorizations, they can make a case, here or at CFD for why these guidelines need to change. Why is this a problem? How does it get in the way? -- Samuel Wantman 22:08, 22 December 2006 (UTC)

Category:Fictional Black African-American DC animated Superheroes with the power to manipulate electricity

I really thought that this one needed to be added to WP:OVERCAT. It is one of my favorite silly categories so far (although "Religious leaders in 1852" and "Red-haired kings" are also very silly). Dr. Submillimeter 23:30, 18 December 2006 (UTC)

That one is quite silly. It's very obvious that Black African-American DC animated Superheroes with the power to manipulate electricity are fictional. -Freekee 19:55, 27 December 2006 (UTC)

Silly category collection

I have started my own silly category collection here. This is a POV collection, although I generally select only categories that are deleted by a near-unanimous vote. It may be useful for gathering ideas on what not to do when creating categories. Dr. Submillimeter 00:32, 19 December 2006 (UTC)

Eponymous categories

Earlier I nominated Category:Joey (TV series) for deletion as an unnecessary eponymous category. Tim! argued that the WP:NOT Wikipedia is not paper policy meant that there is no reason to limit the number of eponymous categories.

I realized that eponymous categories in general might be a good topic of discussion on this talk page. I think most of us would agree that it would be silly for most articles to have their own category (eg. Category:Dandenong Southern Bypass, Category:Ken Duken, a category named after every film and tv show and stage show and book, etc.). At the same time, there are some exceptions involving articles that are particularly notable and/or have articles that span multiple types of topics or genres (eg Category:James Bond, which covers multiple movies, books, actors and other items from different areas of knowledge.)

So in general, it's a bad idea for most articles to have their own unique eponymous category, but under what circumstances should exceptions be made? And would this be something useful to outline in this guideline? Dugwiki 20:24, 20 December 2006 (UTC)

P.S. Feel free to give your feedback on the Category:Joey (TV series) cfd I mentioned. In that case, the category has only one article and one subcategory for episodes, and the episode subcategory is already categorized under Category:Episodes by television series so doesn't need another unique parent. On top of that, anybody seeking articles related to Joey can simply visit the main article and click on the appropriate link; there is no perceivable need to visit "Category:Joey" when you can simply go to "Joey" instead. Dugwiki 20:29, 20 December 2006 (UTC)
My main criteria for categories is "are they useful and is the topic studied". So for an eponymous categories, there should be books about the topic, or academic courses taught for them to be "studied". The "utility" part implies that the category should have some utility that is not found elsewhere. The measure for this, in my way of thinking is "are all the articles related to this topic already cross linked to the eponymous article?" and if they are not "Can a short list of 'See also' entries be added to the article instead of this category?". The other consideration is "Are there articles that are only belong in the eponymous category and nowhere else?". So if the topic is studied, there's a reasonable number of articles, that might be difficult to find from the eponymous article, and/or there are orphans that have no other category to reside in, keep the category. -- Samuel Wantman 22:12, 20 December 2006 (UTC)

To follow up, the result of the cfd for Category:Joey (TV series) was keep (although as a technical matter I think the admin marked it incorrectly - it probably was "no consensus" since I think half the people commenting supported deletion. The result would be the same though.) Given that this category only currently has two items, and that the series isn't any more notable than any other network TV series, I'm wondering which way we're supposed to be categorizing TV shows? Should every TV show with more than one related article have its own category? Or should we be deleting eponymous TV show categories that only have a few articles and that can be safely removed without orphaning articles in the category system?

I also posted this question at WikiProject Television, since all these categories fall under that project. What I'm looking for is guidance on whether I should be creating more eponymous categories for TV shows that don't already have one, or cfd-ing those categories when they appear to be unnecessary. Right now there is an inconsistency in the creation of eponymous categories, both within TV shows and also between TV shows and Films and Books.

Any comments and insight are appreciated, thanks. Dugwiki 16:22, 26 December 2006 (UTC)

Eek! Sorry I missed this one. I would have strongly advocated "delete". I've requested the closing admin to change it to "no consensus" and I'd support more "cfd-ing" of these categories. I think they frustrate users when they entice you away from an article with numerous links to a category with few. They have negative utility. -- Samuel Wantman 21:15, 26 December 2006 (UTC)

I am a little late to this discussion, but I have been addressing the question of eponymous television companies on the cfd page. My original post had to do with re-categorisation, but it ended up touching on whether there should even be separate categories for a producer and his/her production company. It came up in connection with Quinn Martin, but applies to people like Merv Griffen, Reg Grundy, etc. I am generally in favor of having an article for the individual and an article for the corporate entity, although in some instances one or the other may be short. Thoughts?--Vbd 07:27, 17 January 2007 (UTC)

Examples that are beside the point

The project page has the following section

Arbitrary geographical grouping
Examples: Roman Catholic Bishops from Ohio, Quarterbacks from Louisiana, Male models from Dallas, Texas
Avoid subcategorizing items by geographical boundary if that boundary does not have any relevant bearing on the items' other characteristics. For example, quarterbacks' careers are not defined by the specific state that they once lived in (unless they played for a team within that state). However, geographical boundaries are useful for dividing items into regions that are directly related to the items' characteristics (for example, Roman Catholic Bishops of the Diocese of Columbus, Ohio or New Orleans Saints quarterbacks).

The exceptions given to this principle are not geographical units, per se. The 'Diocese of Columbus, Ohio' is an administrative unit of the Catholic church. While it may have defined geographical boundaries, I don't think you would say that someone lives in the Diocese, but you could say that someone is a member of a parish in the Diocese (I'm speaking of usage in the U.S. here).

The 'New Orleans Saints' is definitely not a geographical identifier. It is a team name. At the most, team names represent primary media audience regions.

Of course, I can't think of any pertinent examples to substitute here. Maybe there are no valid exceptions. -- Donald Albury 12:57, 21 December 2006 (UTC)

I agree that the geographical boundaries of the example exception categories are not "real" geographical boundaries. However, the point is perhaps to show how place names should and should not be incorporated into category names. (Part of the purpose of this section is to deal with problematic categories on clergy created by one specific user.) Maybe the language in the section can be rewritten to reflect this. Dr. Submillimeter 14:19, 22 December 2006 (UTC)

Specific cases

I think that this section should be removed. For one thing, it has the greatest possibility of growing exponentially, and at the same time, it's the section most likely to include controversial elements, or elements which may still be under discussion.

What's currently there is either a part of "Performer by performance" (which hasn't been served up for consensus yet, even if we allow a TV/Radio series cast membership exception); or things involving fictional characters (which also doesn't have an overall consensus yet). - jc37 13:06, 21 December 2006 (UTC)

If we cannot give a reason for why something has been removed for overcategorization, or if there is not an established consensus we should probably leave it off the page. I could see having a section to discuss areas that have not reached consensus so that people can be aware of the precedents (both pro and con) and the issues involved. That way, when similar discussions start, we can point people to the section to get background. -- Samuel Wantman 21:45, 21 December 2006 (UTC)
Which sounds like such things belong on a different page (one for discussion, rather than one purporting to be a guideline which has already achieved consensus). I removed them, though obviously someone is welcome to start another page for discussing such topics. - jc37 06:54, 22 December 2006 (UTC)
Discussion is different from a summary of what the issue is and the different sides of the debate and relevant issues. That could still be on the page. I find it very helpful to see guidelines that are as clear about what consensus has not been reached as what has. -- Samuel Wantman 07:02, 22 December 2006 (UTC)
Perhaps, but I think that that is a bit beyond the scope of this page, especially if the wish is to call it a guideline (which suggests stability, and already achieved consensus). As I mentioned above, add those to another page, and perhaps add a link from this page to that one? - jc37 08:43, 22 December 2006 (UTC)
It is possible to explain differences in opinion on a guideline. It is in the spirit of guiding people to tell them there isn't agreement. Here's an example from Wikipedia:Categorization:
In the "vertical" dimension, Wikipedia has traditionally been more frugal, placing articles only in the most specific categories they reasonably fit in. Thus, if there is a Category:American film actors, John Wayne would go there and not in Category:Film actors or Category:American actors. However, there is a school of thought that argues that, because different users may be interested in different categories, and because placing articles in multiple categories takes up minimal additional space, in some cases one should place articles in all the categories that apply. For more about this see Wikipedia:Categorization/Categories and subcategories
Every time we don't spell out what the disagreements are, we have to recreate the wheel. Let's keep the wheel that is half complete. I'm a firm believer in WP:1RR, so I'm not reverting, and I'd like to hear from other contributors to the page. -- Samuel Wantman 09:14, 22 December 2006 (UTC)
  • I think we should list only the specific cases that occur a lot. Otherwise, we can point people to the CFD archives. We've had quite a lot of "film by actor / actor by film" cats on CFD lately. >Radiant< 09:15, 22 December 2006 (UTC)
    • I agree with both sides of the issue here. I agree with Radiant that some of the recurring categories need to be specifically mentioned. However, I also see that consensus could change over time on some of the specific categories (such as comic book character by team categories). In that light, maybe it is worth leaving the specific cases off WP:OCAT for now. However, the individual items in the list could be debated individually here to see if they should be added back to the list. Dr. Submillimeter 14:14, 22 December 2006 (UTC)

Just to follow-up, has there been any further discussion on the actors-by-film and actors-by-tv-series categories? Currently it sounds like there's still some deep division on the by-tv-series topic of which categories should exist, but excluding most actors-by-film categories appeared to have consensus. Right now though all these cfd debates are currently ad hoc with no specific guidelines dealing with actors. It would be nice if the actors-by-film guideline at least appeared somewhere, if not here then possibly on another page linked to by this one (Wikipedia:Overcategorization Specific Cases possibly?) Dugwiki 23:53, 2 January 2007 (UTC)

No potential for growth change

Tim! made a recent change in that section that made sense, so I altered the wording of the provided example. The previous example talked about bands or record labels, but Tim correctly pointed out that records-by-artist is probably an exception to the potential growth rule.

I revised the wording in that section to specify that one exception to the rule is when you have a well accepted categorization scheme, such as Category:Books by author, which subdivides large numbers of articles into categories, some of which might be small. For example, even if an author only has one or two books, it would still make sense to create a category for his books to fit within the Books by author sorting scheme.

Obviously if there is a problem with the wording I used feel free to correct it or comment on it here. Thanks. Dugwiki 18:30, 21 December 2006 (UTC)

I put in the specific example Category:Songs by artist, whose category description emphasizes that all songs should be subcategorized there, regardless of how many songs the artist has recorded (meaning that some subcategories contain only one or two songs). Dugwiki 21:27, 21 December 2006 (UTC)

Caveats

Can we add a little to the opening paragraph that these are guidelines based on precedent, and that they are not codified policy. The consensus for and against some of these categories has changed in the 3 years I've been around here, and I would not be surprised if they change again. I can specifically imagine a big change to this guideline due to technological changes, like for example, if Category intersection ever becomes a reality. -- Samuel Wantman 22:10, 21 December 2006 (UTC)

No further caveat should be necessary because the big "guideline" template box at the very top of the article specifically says "...It is generally accepted among editors and is considered a standard that all users should follow. However, it is not set in stone and should be treated with common sense and the occasional exception. ..." Dugwiki 22:49, 21 December 2006 (UTC)
We don't do precedent. --badlydrawnjeff talk 00:23, 22 December 2006 (UTC)
Actually we do precedent all the time. For instance, just about all of Speedy Deletion was made out of the precedent that such a kind of article appears on AFD a lot and ends up deleted. >Radiant< 09:22, 22 December 2006 (UTC)
Then why is WP:CCC a guideline, since it's supposedly descriptive? d;-) I'll repeat - we don't do precedent. --badlydrawnjeff talk 11:38, 22 December 2006 (UTC)
  • You're wrong, WP:CCC is not a guideline. At any rate its point is obvious - since guidelines are based on consensus, and consensus can change, so too can guidelines change. Indeed, that's why we say they're not set in stone, and why we don't protect them. Check out WP:PPP for an explanation of why we do a lot of things by precedent, mainly to obviate the need redundant debate. >Radiant< 14:23, 22 December 2006 (UTC)
My mistake, it's a policy. Either way, you're making my point. --badlydrawnjeff talk 16:07, 22 December 2006 (UTC)
  • To reiterate, since guidelines are based on consensus, and consensus can change, so too can guidelines change. Indeed, that's why we say they're not set in stone, and why we don't protect them. Check out WP:PPP for an explanation of why we do a lot of things by precedent, mainly to obviate the need redundant debate. >Radiant< 10:14, 24 December 2006 (UTC)
This is a problem. Different people have different ideas of what a guideline is. Some people think that precedent is fundamentally different than rules based on consensus. Radiant and a few others think precedent is consensus - while me and a few others think that consensus is determined by discussion - which can include discussion of precedente - but is not itself simple precendent. Of course, I "fundementally misunderstand wikipedia's policy" according to ArbCom ; ) Fresheneesz 09:46, 24 December 2006 (UTC)
Indeed, you fundamentally misunderstand, and you should read up on WP:POL, WP:HCP and WP:PPP to clear that up. >Radiant< 10:14, 24 December 2006 (UTC)
Despite my lack of understanding... I don't think you understand that precedent is only mentioned in a link to the Notability criteria for books as information on how to include examples. Precedent is therefore not used to create policy - but used to explain (give examples of) policy. Fresheneesz 23:05, 24 December 2006 (UTC)
Precedent is, however, used to create guidelines. >Radiant< 23:56, 24 December 2006 (UTC)
I like how you conveniently leave out any supporting information for that. Why would policy be created in a different way than guidelines? That just doesn't make sense. Fresheneesz 02:46, 26 December 2006 (UTC)
Because policy may contradict current practice (e.g. when they introduced WP:BLP, or WP:FUC formally forbade use of FU images outside article space (leading, amongst other, to the removal of numerous images from templates). Guidelines, however, are not binding (as repeatedly pointed out by yourself), but cannot be established outside current practices. Circeus 05:26, 26 December 2006 (UTC)
So by that definition, the format of essay -> guideline -> policy doesn't work for stuff that goes against current practice. I'm pretty sure that guidelines don't have a set rule that says "guidelines can go against current practice, but policy can". In fact, i'm pretty sure its all based on consensus. Fresheneesz 04:17, 27 December 2006 (UTC)
"A guideline is any page that is: (1) actionable and (2) authorized by consensus." WP:CCC has nothing to do with it right now. If you don't have anything to say about the guideline itself, please stop wikilawyering. If you can't provide arguments to the fact that the page "goes against current practice," please stop wikilawyering. Circeus 04:34, 27 December 2006 (UTC)
Your empty accusations of Wikilawyering are doing nothing to advance your case. As you have yet to demonstrate consensus (as I have raised on this page numerous times), you're not even fulfilling what you believe should happen to make this a guideline, let alone making a cognizant argument as to why this proposal is a good idea. Less attacking other editors, more substance, please. --badlydrawnjeff talk 05:07, 27 December 2006 (UTC)

This argument is still going on... I don't know about everyone else, but I support this guideline because I'm tired of the hundreds of pointless "Fictional X" categories. I see it all the time, someone comes in and starts adding things to "fictional dragons that can drive race cars" and a little time later a bot comes in to remove it due to a CfD. I could start digging around and keep track of all those CfDs if people want proof of consensus, at least for "fictional X" categories. -- Ned Scott 06:02, 27 December 2006 (UTC)

  • Well, it's not really an argument, it's a meta-argument. Nobody is objecting to the page as written. Two people claim there is no consensus, but this is based on what? If there is no consensus for the page, it should be easy to find someone that disagrees with it. Yet despite this page being well-advertised, nobody has been found. So this appears to be another rehash of the argument that "official process was not followed in writing this guideline", which is a vacuous truth since such an official process does not exist in the first place. Wikipedia is not a bureaucracy, so arguments should be based on content rather than procedure. >Radiant< 09:03, 27 December 2006 (UTC)
    • I oppose it as written. --badlydrawnjeff talk 13:27, 27 December 2006 (UTC)
      • The issue is whether the page is accurate, not whether or not you like it. You're treating guidelines as legislation again, but they're not. >Radiant< 13:43, 27 December 2006 (UTC)
        • No, they're not. But if they lack consensus, there's a problem. I don't believe it to be accurate, and I don't see any evidence that it's accurate. My opposition disappears if there's some evidence of consensus from the community that this is how we should be doing things, but you seem against seeking such consensus out. --badlydrawnjeff talk 13:53, 27 December 2006 (UTC)
          • Just about everything on the page is sourced and linked. You can verify the accuracy for yourself. It's not about the opinion of how we should be doing things, but it's about the facts on how we are doing things. Consensus is demonstrated by the fact that it matches actual practice, as well as the fact that this page has been advertised, commented on, reworded per concerns, and generally agreed upon. Consensus is not unanimity, and a consensus that does not include you is nevertheless a consensus. >Radiant< 13:57, 27 December 2006 (UTC)
            • No, it really isn't. If how we're doing things is against community consensus, we should stop doing it. Why do you wish to not involve the community in this discussion? I don't think this is how we actually do things - CFD discussions are low-traffic and we don't do binding decisions anyway. --badlydrawnjeff talk 14:01, 27 December 2006 (UTC)
              • And before saying we're "doing stuff against community consensus," please explains how so?Circeus 14:12, 27 December 2006 (UTC)
              • So you claim that what we're doing in a process based on community consensus is in fact against community consensus. That's an self-contradictory. Binding decisions have nothing to do with it, and guidelines aren't binding in the first place. >Radiant< 14:06, 27 December 2006 (UTC)
                • I claim that community consensus isn't even factored in. You made it a guideline without discussion, and don't care to hear more input, merely shouting louder and louder. I don't buy it, need hasn't been demonstrated, wide community acceptance that this is "how we do things" hasn't been demonstrated. Demonstrate those, and my opposition to you labeling this evaporates. --badlydrawnjeff talk 14:09, 27 December 2006 (UTC)

This whole thread is insanity. Guidelines describe existing practice; a guideline can be prescriptive, but only if the case made for it is so convincing that it changes existing practice. A guideline proposal is rejected if it doesn't describe current practice and doesn't convince people to change current practice. There is some disagreement about whether guideline proposals should be used for only one or the other goal or both or whatever, but debating that somewhat pointless meta-policy argument is not what this talk page is for.

Fresh, contain your comments to the subject of how to use categories. You are stirring up needless drama. Consider this your warning; the next step is going back to Arbcom.

As for everyone else, determining if this is supported by consensus is simple, if not easy. If everyone says "Yes, we should do this" or "Yes, we are already doing this," then it's a guideline. If everyone says, "No, we're not doing this and we shouldn't do it," then it's rejected. Determining which side of the line this falls on may not be easy, but it is a simple line to identify. - A Man In Bl♟ck (conspire | past ops) 09:06, 27 December 2006 (UTC)

Please assume good faith MIB. I'm simply arguing my point. I didn't start the point, so maybe you should ask others to keep their arguments on base. Also, I'd appreciate it if you don't threaten me - but if you want to take this back to ArbCom, be my guest. Fresheneesz 21:39, 27 December 2006 (UTC)
I don't care why you're doing it. You're disrupting this discussion; I'm willing to believe it was unintentional, unless you persist. Contain your comments to the subject of categorizing articles or withdraw from the conversation. If you are attempting to mediate this discussion, I suggest you withdraw instead of continuing to make this situation worse. - A Man In Bl♟ck (conspire | past ops) 05:33, 28 December 2006 (UTC)
If so, then as we currently have 12 semi-disparate entries, we should make that determination for each of them individually... - jc37 09:18, 27 December 2006 (UTC)
And personally, I'm not certain if this isn't better as an essay (after looking over quite a few other essays), while still being very useful as a reference for CfD discussions. - jc37 09:18, 27 December 2006 (UTC)
  • The problem with citing an essay in any process is that people will respond with "you can't cite that, it's only an essay". That's an obvious fallacy, but people will do it anyway. But more importantly, the problem with citing an essay to educate novice users is that they will ignore it since "it's only an essay". And there's a point to that, because anyone can write an essay stating anything - so any page that is significantly more than just "some random editor's opinion" should not be lumped in with the many pages that are, and should called an essay. I'm not particularly happy with the things these terms have become, but that's the way it is at the moment. >Radiant< 09:43, 27 December 2006 (UTC)
Reset tabs

Why discuss existing consensus? May I remind you have brought NO INDICATION WHATSOEVER that anybody else might disagree with the policy, and failed to argue against specifc point in the policy? If this was WP:FAC, by the actionability criteria, your comments would be summarily dismissed. Circeus 14:34, 27 December 2006 (UTC)

But it's not, so I don't have to worry about it. I've made my arguments, why don't you read them before commenting further on what you believe my position to be? --badlydrawnjeff talk 14:35, 27 December 2006 (UTC)
it's clear that without actionable, constructive comments, we can't achieve unanimity.Fortunately, we're looking for consensus. Something that you persistently fail to demonstrate is not reached. Circeus 14:42, 27 December 2006 (UTC)
What you guys need to do is calm down and stop arguing whether hypothetical people may or may not support this as a guideline. Bring in some other people to comment on the page - post it on a noticeboard for comments on its usefulness as a guideline. It might be also a good idea to construct a well worded poll that you can use to gauge the feelings of the community. Fresheneesz 21:36, 27 December 2006 (UTC)
You are being voluntarily misleading: within the last 24h, this has been brought up at WP:VPP and WP:RFC/POLICIES. And any further suggestion of a poll will lead to a request from me to put you on probation. Circeus 22:26, 27 December 2006 (UTC)
Circeus, I think we can work this out without judgements as to whether anybody's being "voluntarily misleading" or threats of probation. Your reply to the content - pointing out that it has been brought up at VPP and RFC/POLICIES is sufficient, surely? I agree that a poll is probably a bad idea in this case, as those tend to create polarization and conflict where none existed before. -GTBacchus(talk) 22:31, 27 December 2006 (UTC)
Freshneesz attitude so far in this discussion has been nearing disruption (see [1], and [2], not to mention his comment as to having "no opinion", if so, what is he doing here in the first place?), and he has been warned by the ArbCom over that. His appearance here came out of nowhere and looks more like a need to argue with Radiant than anything else. Circeus 23:17, 27 December 2006 (UTC)
So the question becomes, how do we respond to "near disruption". I wouldn't suggest that claims about people's motives or threats are the most effective way, but your mileage may vary. I'll leave you to your best judgement. -GTBacchus(talk) 23:40, 27 December 2006 (UTC)
After Freshneesz's non-notability debacle (even though I was not involved in it, it was a pretty ridiculous thing overall), I'm inclined to little tolerance of any suggestions of polling from him. And his arguing here does not appear to have been in the spirit of solving the debate (Refering here to his "no opinion comment"). Circeus 01:02, 28 December 2006 (UTC)

Unless someone else independently suggests a poll, let's put this thread to rest. Clearly outside opinions on what this is (a description of existing practice, a good idea for changing existing practice that can find consensus, or just a bad idea) are more useful than arguing about parliamentary nonsense. - A Man In Bl♟ck (conspire | past ops) 05:33, 28 December 2006 (UTC)

Wow. I was definately mediating there. Way to blow up on me. Btw, I looked at the WP:VPP and WP:RFC/POLICIES links you gave - and the *only* outside opinion given on VPP was Francis Schonken - who does not like the idea "imho". Quite bluntly, this needs much more outside commment - much more than 1 dissenter that is. Fresheneesz 11:16, 28 December 2006 (UTC)
I am speechless. Considering your background With radiant and the way you barged in to agree outright with Jeff, nobody in their right mind would ever believe. You are also compeltely misrepresenting the point of WP:RFC/WP:VPP. We want to get comments HERE not on the other side of the project! Circeus 14:40, 28 December 2006 (UTC)
  • "Fresheneesz has shown by his comments that he fundamentally misunderstands how Wikipedia treats policy, and how it is created. He has stated that guidelines need not be reflections of common practice, and that Wikipedia resolves discussions through the use of voting." Those who forget history are doomed to repeat it. >Radiant< 14:58, 28 December 2006 (UTC)
Radiant, please stop parading around those quotes. You've had your fun. While you and those arbcom people might have a different opinion of wikipedia than I do - I share my wikiphilosophy with a large group of people on here - I humbly ask you to quit throwing personal attacks my way.
To Circeus: What does my background with radiant have *ANYTHING* to do with this discussion? I don't even know jeff - i just agree with him. YES *WE WANT COMMENTS HERE* - I agree wholeheartedly. So where the hell are those comments? We cannot base consensus off of 5 people, even if those are the only ones that show up at this page. A consensus of 5 is the same thing as "no consensus". Fresheneesz 00:11, 30 December 2006 (UTC)
I once proposed quite the opposite at the Village Pump. That we require at least a minimum quorum of 5 WIkipedians in order to determine consensus. While there were several who agreed that at times 3:1 can bully the 1, there was no consensus for such an implementation, since there are different venues for consensus in the many processes of Wikipedia, and because of that many target numbers for consensus. All of that aside, your statement is fallacious on face value: "A consensus of 5 is the same thing as "no consensus". " - If there is a consensus of "some number" (5 or 5 million, is irrelevant), then there is a consensus. Whether or not we have consensus on this page, is, I think, more what you're trying to determine. (And of course, remember that quantity/ratio of votes does not necessarily equal consensus or no consensus). - jc37 11:00, 30 December 2006 (UTC)
My point was that we need more people to comment on this page. Fresheneesz 22:23, 30 December 2006 (UTC)
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Overcategorization/Archive_1&oldid=1137431305"