User:Sidatio/Conversations/On list guidelines

From Wikipedia, the free encyclopedia

This page exists to discuss a proposed policy guideline on lists. For more information on the background of this project, see User:Sidatio/Conversations/On_list_guidelines#Background.

Contents 1 Background 2 TO THE EDITING COMMUNITY AT LARGE: 2.1 Arbitrary break 2.2 Proposal for a very short guideline/policy statement 2.3 On the contrary 3 Day 2 3.1 Categories need to be a tool, not a source of navigation frustration 3.2 Looking good! 4 Maybe I don't get it... 5 PRELIMINARY GUIDELINE IS UP FOR REVIEW 6 Day 3 7 Day 4 7.1 About "joins" of list inclusion criteria 7.2 The problem is lack of clear inclusion criteria for stand-alone lists 8 Day 5 9 Under new management 10 test driving on lists of people 11 test driving on other lists 12 Evaluate WP:Not? 13 Alternate proposal: clarification and alteration of up to three policies 14 Question 14.1 Ah, this is what it was!

[edit] Background

This discussion started at the Village Pump in an attempt to get an idea of what current consensus was in regards to WP:LIST guidelines - namely, do the current guidelines offer enough guidance during list creation or AfD discussions? Recent AfDs involving lists seemed to reflect this confusion and did little to establish consensus; Wikipedia:Articles for deletion/List of Indian women failed to pass its AfD, while a similar list, Wikipedia:Articles for deletion/List of Iranian women failed to garner any consensus. Further, lists seemed to be nominated more and more based on criteria outlined in WP:LISTCRUFT, which is an essay - not a guideline or policy. Meanwhile, a search through AfD archives showed me that some of the most heated, contentious AfD debates took place over lists.

With all of this in mind, I went to the Village Pump with the idea of creating a new policy or guideline governing the viability of lists, using WP:LISTCRUFT as a point of reference. It soon became clear that this wasn't the route to take; rather, the aim should be in the modification of WP:LIST to make the guidelines more clear. It also became clear that much of WP:LISTCRUFT was either already applied in other article guidelines, were too vague as written, or simply didn't apply. So far, though, consensus seems to indicate that there needs to be better definition on the subject.

The conversation has been copied from its original starting point at Village Pump, and has been moved because of its size to this page, where the discussion can continue unabated. Participants include deletionists, inclusionists, exclusionists, guys named Sterling, some of Wikipedia's most experienced editors, and some of Wikipedia's newest. Assuming good faith and civility are requested of the participants, but that really should go without saying. Suggest anything you can think of in relation to the topic - hopefully, we can all help work toward a guideline that better defines inclusion criteria, and, as a result, helps to defuse some of these heated AfDs and save some lists that might otherwise be lost (or delete some lists that might otherwise be saved).

And now, the conversation:

[edit] TO THE EDITING COMMUNITY AT LARGE:

I am writing this in the attempt to find consensus for the need of a the expansion of policy or guidelines for the creation and maintenance of lists - specifically, WP:LIST#Lists content or perhaps its own section on that guideline.

From what I can tell, WP:LIST is vague at best in its guidelines for lists in general, and their viability in particular. Also, Wikipedia:Deletion policy#Reasons for deletion addresses some of the reasons an article can be nominated for deletion, but it doesn't address any of the reasons lists are commonly deleted. Further, there is only one entry under Wikipedia:Articles for deletion/Precedents#Lists. This entry in and of itself is also vague - under what circumstances are they often kept? How often is "often"?

Recently, it was argued on Wikipedia:Articles for deletion/List of chefs (and rightly so) that criteria found on WP:LISTCRUFT are not, in fact, policy. Rather, it is an essay on the topic of list criteria. However, Wikipedia:Articles for deletion/List of Europeans, Wikipedia:Articles for deletion/List of African Americans (3rd nomination), Wikipedia:Articles for deletion/List of Indian women, and Wikipedia:Articles for deletion/List of living philosophers and academics of philosophy are just a few lists that were found by consensus to be lacking based on several criteria outlined by WP:LISTCRUFT. There were also some arguments made from the Wikipedia:Overlistification essay that seemed to be accepted by consensus.

In sum, there seems to be at least SOME consensus by seasoned and "green" editors alike as to the validity of WP:LISTCRUFT and Wikipedia:Overlistification as guidelines for list articles. Likewise, there is also opposition to the article in whole or in part, much like with almost any criteria in use today on Wikipedia. So, here I am with an attempt to find middle ground and, hopefully, get some kind of formal guideline formulated to address the issue.

My arguments for the creation expansion of a policy/guideline are as follows:

• Maintainability. As found with Wikipedia:Articles for deletion/List of people by name, as well as the provided AfDs above, there is consensus among community members for using maintainability as a criterion for keeping or deleting a list. By nature, lists that are this open-ended would require near-Herculean efforts by a fairly large number of editors to properly maintain to encyclopedic standards. In cases like these, there are only two reasonable outcomes - the lists take the time and effort of a disproportionate number of editors to maintain and expand the list to the point that it negatively affects more and more articles unrelated to said list, or the list itself falls into disrepair and becomes obsolete as a result of too few editors maintaining and expanding the list. Either scenario would reflect poorly on the quality and accuracy of Wikipedia. A guideline or policy that further expounds on criteria for the maintainability and viability of a list article alone would go a long way to preventing issues like these - and in many cases, already has without the specific definition of such a guideline based on community consensus and precedent.

• Overlistification. Some lists are just too narrow in scope to be useful - for example, Wikipedia:Articles for deletion/List of songs by African, Asian, Caribbean and Latin American artists which reached number one on the Hot 100 (US) and Wikipedia:Articles for deletion/List of software projects whose name is a term offensive to many people with disabilities. A guideline or policy stating an acceptable scope for a list would be very helpful in creating lists of articles that could be considered detailed, and would also help to better establish consensus on how narrow is too narrow.

• There's nothing to properly cite: As it stands, the best articles on the subject (that I have seen, anyway) are mere essays. Without codification into a policy or guideline, there's really nothing to concretely use for or against the creation or retention of almost any list on Wikipedia. As such, lists themselves fall into a policy "gray area" that can be resolved with the implementation of a policy or guideline on the subject.

To this end, I endorse the implementation of the guidelines contained in WP:LISTCRUFT as an expansion of official guideline or policy governing lists to remedy the above. In sum, a list should not be created or kept if:

1. The list was created just for the sake of having such a list
2. The list is of interest to a very limited number of people
3. The list is a violation of Wikipedia is not an indiscriminate collection of information (which, in and of itself, is already viable criteria anyway)
4. The content is unverifiable or the underlying concept is non-notable (again, already viable anyway)
5. The list cannot be expanded beyond a handful of terms (this may need discussion - how many is a "handful"?)
6. The list is unlimited and/or unmaintainable
7. The list has no content beyond links to other articles, so would be better implemented as a (self-maintaining) category (subject to debate - on one hand, it can be seen as an extension of WP:NOT#DIR; on the other, it severely reduces the visibility of otherwise notable topics)
8. The list is unencyclopaedic, i.e. it would not be expected to be included in an encyclopaedia.

That's the gist of my proposal. I will post links to this proposal to relevant AfD discussions, as well as inform some of the more prolific members of Wikipedia (based on List of Wikipedians by number of edits - probably the top 20 or so, provided they are active) of this debate.

Sidatio 15:18, 13 August 2007 (UTC)

• The first criterion is just an insult to anyone who worked on the list, how are you supposed to figure what their motives are anyway? "The list is unlimited" applies to rather a lot of lists which have wide acceptance, such as list of Poles and List of poets. Open-ended lists can be divided into sections and split when they get larger, per Wikipedia:Lists_(stand-alone_lists)#Appropriate_topics_for_lists. Most of the rest is redundant with WP:NOT. Kappa 16:12, 13 August 2007 (UTC)

Request for clarification Could you kindly source and expound upon your position, please? Specifically, Most of the rest is redundant with WP:NOT - what parts of the above are redundant to which sections of WP:NOT? Thanks! Sidatio 16:28, 13 August 2007 (UTC)

Item #3 is obviously redundant with WP:NOT. #2 is likely redundant with WP:N. #4 is redundant with WP:V and WP:N. #8 is redundant with "not indiscriminate" again. #7 is not currently redundant but it was until recently. For a list to be truly redundant it must lack content, structure, references, sometimes even piping is worthwhile, because it turns "List of x in y" into "y", as List of schools by country does. Kappa 16:51, 13 August 2007 (UTC)

I agree with Kappa in that some of these criteria do seem redundant with the existence of other WP:PAG, but I also think as Moonriddengirl has stated below that the main WP:LIST page needs to be updated with community consensus, which we can find here and then do per WP:BOLD (other users can implement their own changes to these criteria on WP:LIST once we've added them). Anyway, gotta go to work now.-h i s ^{s p a c e} _{r e s e a r c h} 08:08, 14 August 2007 (UTC)

• Having been involved in several AfDs for lists lately (some that have been deleted, some kept), I think the creation of a clear policy is an excellent idea. In conjunction, I think there needs to be clarification on the usage of categories versus the usage of lists, since some AfDs on lists have led to proposals to categorize, while some CfDs on categories have led to proposals to listify. There's circular arguments going on. Clear policy would be helpful to primary editors as well as to editors engaging in debate over the value of specific lists & categories. Kappa, WP:NOT doesn't seem to me in its current form to provide enough guidance on this subject, since I've seen it both argued in AfDs and ignored. I believe there is genuine confusion on both sides of the fence as to how to handle lists, and I think that whether listcruft is adapted or some other guidelines devised, more guidance could be very helpful. --Moonriddengirl 16:26, 13 August 2007 (UTC)

• As far as AfD is concerned, lists are real simple. 1. Does the list include an unambiguous statement of membership criteria based on definitions made by reputable sources? 2. Is the list adhering to that membership criterion, including significant efforts to use reliable source material for EACH list entry? Generally, if the answers are no, then delete. There may be exceptions. In the end, if consensus decides to Keep the list, then that probably is what is going to happen no matter how much text is used in a policy or how much "guidance" is provided. Also, I agree with Kappa, there is no need for more policy as current policy covers the proposal. -- Jreferee ^(Talk) 16:52, 13 August 2007 (UTC)

Respectfully, I disagree with the "lists are real simple" for AfD assessment. :/ List of Indian women closed with delete; the obviously very similar List of Iranian women closed with no consensus, which defaults to keep. Clearer policy might make at least some consistency possible. --Moonriddengirl 17:03, 13 August 2007 (UTC)

• Here's my opinion on these criteria one by one.

1. The list was created just for the sake of having such a list

How would you even judge this? Not an objective critereon, difficult to apply.

2. The list is of interest to a very limited number of people

Same as #1.

3. The list is a violation of Wikipedia is not an indiscriminate collection of information (which, in and of itself, is already viable criteria anyway)

Valid, if redundant.

4. The content is unverifiable or the underlying concept is non-notable (again, already viable anyway)

Valid, if redundant.

5. The list cannot be expanded beyond a handful of terms (this may need discussion - how many is a "handful"?)

I don't see why this is an issue.

6. The list is unlimited and/or unmaintainable

Valid.

7. The list has no content beyond links to other articles, so would be better implemented as a (self-maintaining) category (subject to debate - on one hand, it can be seen as an extension of WP:NOT#DIR; on the other, it severely reduces the visibility of otherwise notable topics)

If the list provides no structure and could be duplicated by a category, certainly. But there are many things that a list can do better than a category.

8. The list is unencyclopaedic, i.e. it would not be expected to be included in an encyclopaedia.

Vague. Seems to run counter to a number of Wikipedia policies and guidelines.

-Chunky Rice 16:53, 13 August 2007 (UTC)

Comment: By and large, I agree with Chunky Rice's assessment of the WP:LISTCRUFT criteria, especially number 6. However, number 6 currently isn't part of any criteria for creation or retention of lists that I have seen, other than essays that aren't true criteria by definition. However, this is one of the most cited reasons for deleting a list, as outlined in my examples above. (I can provide more examples if need be.) As to numbers 3 and 4, yes, they are accepted criteria - I only listed them in summary of WP:LISTCRUFT. As far as number 7 goes, I can see that turning into a whole 'nother debate based on the outcome of Wikipedia:Articles for deletion/List of Iranian women. And yes, number 8 is too vague. Are there any ideas on whether or not this can be narrowed or better defined, or should it be scrapped? Personally, I think it's important to an extent, but it DOES seem to be contradictory to several accepted policies.

In regards to Jreferee's comment: The problem there is that the two-part criteria you outlined seems to be not only vague, but flawed. There are several lists failing that criteria that could be deleted, but a fair amount that fail to measure up to those standards can also be restructured and saved. Further, it doesn't seem to address viability or maintainability, which I feel are important for the reasons I outlined. Sidatio 17:08, 13 August 2007 (UTC)

• This belongs on Wikipedia talk:Lists: WP:LIST is a guideline, and as such is intended to represent our current consensus on lists, to be taken into account in afds. We need to keep the number of actual policies at a minimum. Everything on WP:LIST should be a reasonable corollary of Wikipedia:Notability, WP:ATT and WP:ENC. Policy is "it needs to be notable. it needs to be verifiable". A guideline tells you "ok, for lists, this translates as follows ...". What is the point of having guidelines if people think they can just ignore them on afd? If some list is in blatant violation of WP:LIST, ask editors to seek for a new wiki-wide consensus on lists first. That said, I am all for a more stringent guideline on lists. This means, you want to improve WP:LIST, you don't want to introduce new policies. --dab (𒁳) 17:18, 13 August 2007 (UTC)

• Makes sense - I was under the impression that WP:LIST was more of a style guideline, rather than a guideline that offered boundaries on the creation/retention of lists. Any consensus reached here, however, could definitely serve to expand WP:LIST#Lists content. However, since conversation has already begun in earnest on this page (and also because I put out SEVERAL links to this particular section), I don't see the harm in keeping it here. I think a valid compromise in this situation would be to amend the proposal to state that it wants to expand WP:LIST guidelines on the subject, and leave a link on the Wikipedia talk:Lists page to point to this discussion. Any dissent? Sidatio 17:31, 13 August 2007 (UTC)

That's probably the best way to handle this if we're not going to transfer it over. Silver seren 18:54, 13 August 2007 (UTC)

• Like Chunky Rice, it's easier to handle the components one by one.

1. The list was created just for the sake of having such a list

The criteria for this definitely needs to be more specified because you could say that many lists on Wikipedia are just for having a list. Maybe more on the lines of "the sake of having a list that isn't necessary because there is already a main page on the subject that has the information on it"? My use of words is not that good, I know, but something in those terms should work.

2. The list is of interest to a very limited number of people

The term "limited" definitely needs to be addressed. How is this term going to be defined? There is no actual way of leaning how many people are interested in it and you may end up with most of those people being on Wikipedia and defending it. Obviously, it can't be defined by the number of Wikipedians defending it. Maybe this should go more with the term of "narrow" instead?

3. The list is a violation of Wikipedia is not an indiscriminate collection of information (which, in and of itself, is already viable criteria anyway)

Completely valid, no need of discussion.

4. The content is unverifiable or the underlying concept is non-notable (again, already viable anyway)

This goes with any article on Wikipedia, so yeah.

5. The list cannot be expanded beyond a handful of terms (this may need discussion - how many is a "handful"?)

A handful, in my opinion, is either 5 or 10. Take your pick. Though I don't believe that the number of terms is important if the information is notable and sourced correctly. More than 1, obviously, to make a list, but beyond that...you go into vague territory.

6. The list is unlimited and/or unmaintainable

I agree with this, but also have a question. If there is an imaginary group with imaginary members that is notable and can be sourced, but this group is still accepting members for a certain amount of time, is it still okay to create a list? If yes, then how long is this length of time allowed to be before it goes into the realm of unmaintainable? If no, then the list should be made after the amount is time is closed, I would assume?

7. The list has no content beyond links to other articles, so would be better implemented as a (self-maintaining) category (subject to debate - on one hand, it can be seen as an extension of WP:NOT#DIR; on the other, it severely reduces the visibility of otherwise notable topics)

I agree with Rice on this one. Lists are better in separating and describing than categories, while categories are better at keeping up to date on information. Definitely debatable...I won't be the one to debate it though.

8. The list is unencyclopaedic, i.e. it would not be expected to be included in an encyclopaedia.

This needs to be more specific. I'm not sure exactly what this entails.

And thats what I got. Comments?Silver seren 18:54, 13 August 2007 (UTC)

• • Preliminary comments --I'll comment in detail tonight. But I have two immediate suggestion--there is enough interest to move it to a discussion page of its own, perhaps as a centralized discussion. Second, there is a fundamental division about what is relevant content, and the attempt to produce rules is futile unless there is some agreement or compromise on the basic issue. Essentially, about half think the guidelines as LISTCRUFT too restrictive, and the other half want to enforce them more strictly. DGG (talk) 18:56, 13 August 2007 (UTC)

• I don't see this at all. Lists have made huge contributions here, over the years, both in terms of navigation round the site, and by providing redlinks to work on. That is, they contribute to WP as hypertext, and they promote growth. While categories arguably can do the first, they certainly are quite incapable of doing the second. To put it all another way, some lists are like scaffolding, which can be taken down when the building is over, but cannot just be dispensed with. The idea that lists are just another kind of article 'content' strikes me as deficient in understanding of how WP works. And the list of criteria as a charter for more deletionist nonsense. Some lists don't need to be here, and others are very useful, but on the whole retaining lists that aren't actually silly does little harm and helps navigation. Charles Matthews 18:56, 13 August 2007 (UTC)

But don't you agree that there need to be more specific guidelines for lists in general? If you don't agree with the ones shown here, what are your alternatives?Silver seren 19:17, 13 August 2007 (UTC)

Don't you think a set of criteria called 'listcruft' gets off on the wrong foot? 'Cruft' is an offensive term to apply to the work of others. The fact that some people seem to think it is just part of the jargon is their problem. I say that a list on the site that provokes the creation of some good new articles is a good list to have around, whether or not it meets someone's fairly arbitrary 'criteria'. This is not a beauty contest. It is, often enough, a question of opening up an area of the site that is so far undeveloped. Such lists may not look like much, but can contain useful redlinks. We don't need 'more specific guidelines' for just everything; a list that contributes to WP's encyclopedic mission is a good list. Charles Matthews 06:56, 14 August 2007 (UTC)

Comment: Okay, first off, let's not use labels. For the purposes of discussion, we're not "deletionists" "inclusionists" "exclusionists" or "people named Stan". (Unless, of course, your name IS Stan.) If there's one thing that's evident from the multitude of comments here, it's that there's a need for more guidance on the subject in order to avoid many of the issues being raised here in the future. Civil discourse is the only way to arrive to a reasonable conclusion.

As to moving this conversation to a dedicated page, I agree with DGG - especially since conversations here are auto-archived and may be buried. If there's no suitable place to move this discussion, I'll happily create a subpage on my user section to continue this conversation, with notes to redirect anyone interested in contributing who happens to be "tardy to the party", as it were. Would one of the more experienced administrators have any recommendations as to where to move this discussion if my option is not suitable? —Preceding unsigned comment added by Sidatio (talk • contribs) 15:26, 13 August 2007

Thanks to Moonriddengirl for catching my lack of signature. :-) Sidatio 19:30, 13 August 2007 (UTC)

Oh darn, I want my name to be Stan now...I could go get that legalized. Silver seren 19:44, 13 August 2007 (UTC)

We need to go back to the basic reasons for having and not having articles. Basically we should have them if they are useful/interesting. Useful to people who use Wikipedia - and might be reasonably expected to look for list that they would find in a paper encyclopedia, but also other lists since WP has so much more scope than any previous 'pedia. It is a fallacious argument, often used, that a list in general should only include items of importance enough to warrent (or have) a WP article. Nontheless we can make guidelines (pace the "database" upgrade), which should include IMHO:

1. The collection of things (partially) listed should have a discernable identity, and be notable.
2. The list should state the criteria for inclusion, which should be capable of being citable in priciple.

Some of the ideas above are also usable.

We should also address the question of categories. Currently some categories are part of the WP encyclopedia and others are adminstrative or maintainence. Further many categoriesd are not easily navigable, and leaf nodes have only 1 or 2 members. We should look at dynamic ways of building lists from categories. Rich Farmbrough, 22:21 13 August 2007 (GMT).

• Comment On categories - that's probably going to be a whole 'nother argument based on what happened at Wikipedia:Articles for deletion/List of Iranian women. We're probably better off focusing on one issue at a time, lest we overload ourselves! ;-)

Also, I'm going to move this to the Conversations section of my user space when I get home tonight. Fear not - I'll leave behind a link to the discussion. In the meanwhile, we've got some GREAT discourse going here. I feel confident we can make some real headway in clearing up some of the "fuzziness" of this particular topic. Sidatio 22:22, 13 August 2007 (UTC)

• Okay, I invited a BUNCH of other people into this conversation to see what consensus is. Further, I've discovered that, once consensus is reached one way or another, we can implement the change ourselves if we like. For those of you who chose to participate, I sincerely hope you help me see this one through, one way or another. I've seen other discussions on this topic die due to lack of interest, and I'm determined to avoid that outcome here. We have the opportunity to effect real change here. I appreciate all of you helping to make that happen. Sidatio 01:50, 14 August 2007 (UTC)

I won't be leaving, but you know that. *whistles and waits for more people* Silver seren 01:52, 14 August 2007 (UTC)

• Comment by Mister.Manticre

Ok, here's my comments...which I'll do by number as well.

1. The list was created just for the sake of having such a list

Not the best way to phrase it. I would suggest something like "The list offers more than the equivalent category would" though how to handle the redlink issue, I don't know.

2. The list is of interest to a very limited number of people

This would lead to arguments.

3. The list is a violation of Wikipedia is not an indiscriminate collection of information (which, in and of itself, is already viable criteria anyway)

Here, I would be specific and describe the indiscriminate things.

4. The content is unverifiable or the underlying concept is non-notable (again, already viable anyway)

No objection here.

5. The list cannot be expanded beyond a handful of terms (this may need discussion - how many is a "handful"?)

In this case, I would offer the suggestion of moving into another list of broader scope. At least in some cases. For example, the list of winners of multiple Nobel prizes is small. However, that in itself might be a valid part of the list of nobel prize winners.

6. The list is unlimited and/or unmaintainable

I would also point out that this might have some issues (people will argue over what's unlimited or unmaintainable), and that in some cases, applying a selective criteria would resolve it.

7. The list has no content beyond links to other articles, so would be better implemented as a (self-maintaining) category (subject to debate - on one hand, it can be seen as an extension of WP:NOT#DIR; on the other, it severely reduces the visibility of otherwise notable topics)

This brings up the redlink issue, and also the fact that some lists could be improved by adding basic facts. Take List of assault rifles. Pretty bare list. But if we were to add things like the country of origin, date, type of bullet fired, it'd be widely expanded.

8. The list is unencyclopaedic, i.e. it would not be expected to be included in an encyclopaedia.

Probably something to be defined first.

Anyway, that's what I think for right now, I may have more thoughts later. I do think some examples would be beneficial at least for working with though. FrozenPurpleCube 02:03, 14 August 2007 (UTC)

Any list policy has to take into account the scope of potential additions to the list. I just deleted Kappa's deletion notice on "List of Latvians." Yes, it is a list of names. No, it is patently ridiculous to have a policy that automatically assumes said list will grow to unmanageable proportions. If it grows to the point of requiring aforesaid massive intervention, quite a number of editors would be happy to contribute to the conversation of moving toward creating a category and how that should be organized. In the meantime, insertions of delete unless you respond to this deletion notice for not conforming to a policy which contextually does not apply is not constructive.
    I would suggest a policy that suggests positive steps in making the unmanageable manageable, not merely deleting the unmanageable or the projected to maybe be at some point unmanageable. Deletion implies the information contained has no use. I would emphatically suggest the editors who thing AfD is an actual answer to a list content problem reconsider their position. You don't chop someone's head of to stop their headache, do you? —  Pēters J. Vecrumba 02:11, 14 August 2007 (UTC)
P.S. I came across the aforesaid deletion notice after spending the first Wiki-evening in a very long time providing some constructive information on a topic rather than arguing over something. Then I found this, sorry, Wikidiocy. Other than a basic criterion of notability (e.g., "list of sizes of household dust particles"), any other policy regarding lists should focus on improvement, not excuses for deletion. —  Pēters J. Vecrumba 02:17, 14 August 2007 (UTC)

[edit] Arbitrary break

Thanks for the invite to comment. First, let me preface by saying that I don't like the word "Listcruft" maybe because I'm a humanities guy not a computer guy, but I guess anyone who ventures into the AFD world becomes accustomed to the term and what it tends to mean. That said, I'd recommend these as the most often criticisms of lists, copied from above and edited:

1. The list was created just for the sake of having such a list

• This is I think a recapitulation of several of the others, but seems OK

The list is of interest to a very limited number of people

• I would strike this criteria. It has been criticized, and rightly so: a list of people or things that are obscure but notable shouldn't be deleted due to how often someone would really be interested in it List of Governors of Tokat not pretty, but seemingly acceptable; List of coralline algae species in the British Isles ditto; you get the point. This seems the anti-poisiton to WP:USEFUL.

3.The list is a violation of Wikipedia is not an indiscriminate collection of information

• This is often used improperly, the key word is "indiscriminate" which I would in fact place in bold.

4.The content is unverifiable or the underlying concept is non-notable

• This is a good keeper

5.The list cannot be expanded beyond a handful of terms

• Probably OK, but handful is in the eye of the beholder (as you note)

6.The list is unlimited and/or unmaintainable

• This is one of my favorites: this is what cats are best for. I would reword slightly: "The list is potentially extremely large and/or maintainable" For I have seen people say: it's not too big yet and if it gets there we'll deal with it. Well, let's not have people's efforts wasted building something that will only be deleted if they are successful.
  • I would suggest coming up with a manual of style for what to do when a list becomes too big and borders on becoming a mere amorphous mass. The assumption that you delete what you don't know how to manage is no excuse for not figuring out how to manage it properly. —  Pēters J. Vecrumba 02:23, 14 August 2007 (UTC)

The list has no content beyond links to other articles, so would be better implemented as a (self-maintaining) category

• I have been persuaded that there is value in having the red links around: 1) it invites creation of the articles; 2) it signifies to people that the article is missing/needed; and (most importantly) 3) it fixes the title of the article to WP norms so we don't have 2 of them floating around.

8. The list is unencyclopaedic, i.e. it would not be expected to be included in an encyclopaedia.

• This is in essence a repetion of the non-notable concept above

Then I would a few others:

9. The list is subjective or original research

• When there is no objective criteria for what is in or out and/or verification is possible but difficult, example: List of bad Kings of England, List of good Kings of England, List of Kings of England who would have liked rap music

10. The list is a synthesis in violation of WP:SYNTH

• Certain lists are just POV pushes, possible examples: List of vegan centenarians, List of pro-war people who avoided military service, List of toxic chemicals used in XYZ Corp.'s products, List of actors who have won X Golden Globes but no Oscars, etc....

11. The list is objective but it's someone else's subjectivity

• Many of these end up being copyvios, but some aren't: List of XYZ's top 100 ABCs, List of XYZ travel guide's top destinations, List of Oprah's recommended books

12. The list is an attack list

• Yes they exist and are mostly violative of WP:SYNTH but not always: List of vice presidents who shot somebody (a category of similar name did exist for a while), List of celebrities with drunk driving arrests, List of vegans who wear fur

My $0.02 and I hope all these possible examples turn red upon submission. ;-) Carlossuarez46 02:17, 14 August 2007 (UTC)

Whew, we can breathe easier now. They are all red. ^_^ It would have been really weird otherwise. Silver seren 02:22, 14 August 2007 (UTC)

Oh...and I agree. Silver seren 02:22, 14 August 2007 (UTC)

Anybody want to check the logs to see if the lists *did* exist? FrozenPurpleCube 03:51, 14 August 2007 (UTC)

• Comment by Badagnani - The subtext of this proposal seems to be "I see a lot of content I don't feel is important so I'd like to delete it." I've seen too many lists and list-type articles that were of great significance deleted entirely by the "AFD page regulars" who had no idea about the subject they were deleting. Content dedicated editors had worked on for years, in their areas of expertise, and sharing this information with the world, was, and is gone forever. WP is self-policing in the way that if the editors active in specific subject areas find a particular list to be irrelevant or useless, they can discuss and modify or delete content. But I do not like this sweeping proposal. The articles mentioned (List of Indian women) are obviously too big in scope for WP (though "List of Indian playback singers" or "List of Indian sitarists" would probably not be too big.) This sort of thing should be evaluated on a case-by-case basis and what is considered "junk" to one editor may be considered very valuable by others. Badagnani 02:23, 14 August 2007 (UTC)

• Comment I think you have made this far more complicated than it needs to be. Unless the article is a disambiguation page, being a list should qualify for speedy deletion. I stumbled across http://en.wikipedia.org/wiki/Category:Lists_of_schools_by_country today, and became immediately daunted by the task ahead.Kww 02:27, 14 August 2007 (UTC)
  • Comment Your oversimplification assumes lists have no informative value. If you're daunted by something, figure out how to manage it. Deletion is not management. It's the intellectually cowardly way out. —  Pēters J. Vecrumba 02:50, 14 August 2007 (UTC)

Comment: Well, so much for assuming good faith and civility, eh?

• Well, I think the previous statement is a bit uncivil, I do think that the statement that lists should qualify for speedy deletion is a very bad idea, and indicates to me a lack of awareness of the value of lists. If lists were *that* bad, we wouldn't have Featured Lists. In any case, I strongly advise against making a deletion for that category, as I feel it'd be a waste of time. FrozenPurpleCube 03:51, 14 August 2007 (UTC)

First and foremost, let me clarify: We're not looking to find ways to automatically delete ANYTHING here. It's simply a discussion to try and find consensus, and throwing around labels and names WILL NOT ACCOMPLISH ANYTHING. There are probably as many opinions on the subject as there are people, and we're not going to get anywhere by assuming the worst of people. It's not about the effort someone may or may not have put into a list, it's not about inclusionism or exclusionism. It's about civil discourse.

• Comment Unfortunately my participation in this "conversation" was prompted when this conversation manifested itself as a deletion notification for a list "violating" said discussed Wiki list "standard." (Justification pointed to conversation ultimately directed here.) If the issue is managing voluminous and/or complex information, let's address it positively, not summarily decide, e.g., "unless a disambiguation list, speedy delete", what to delete. Deletion appears to be the only editorial action being suggested here, other than a "keep" which is no editorial action. What I consider "uncivil" is editors picking up on the potential for contributing positively with deletion notices (which mean people's good work destroyed permanently in five days unless responded to). If that is not the intent, then the nature of the conversation here needs to change. —  Pēters J. Vecrumba 04:50, 14 August 2007 (UTC)

Secondly, I don't necessarily like the term "listcruft" myself; that just happens to be the name of one of the essays on the topic. (Personally, I like Wooty coining the term "fingercruft", but I digress.) It's imperative in a discussion like this to assume good faith - I personally am not out on some deletion rampage, nor do I believe there's room for everything and their mother (even my mother). This is just a discussion to try and find consensus on what is (strangely) a hot-button topic.

Thirdly, a quick response to some of the comments I've seen. I'd go more in-depth, but it's bedtime for me. (I figured I'd pop on real fast and try to preemptively defuse some of the angrier comments after checking this on my cell, though!)

• To Badagnani: I apologize if you thought my intention was to come up with reasons to delete content. I hope I've cleared that up for you. Also, why would the list of Indian women have been too big, yet the List of Iranian women, being of the exact same scope, ended up defaulting to a no consensus? That's why I opened this conversation - to find that consensus. :-)

• To Pēters J. Vecrumba: I appreciate your additions and your passion, even if it is a bit strongly-worded. Could you kindly dial it down a notch, please? I'd really appreciate it. I know there are strong feelings on either side, but I think we can all agree that civil discourse is the best approach here, right?

• To FrozenPurpleCube and Carlossuarez: I like the "breakdown" approach you've taken here (this also applies to the original contributors who used that approach - I think it's VERY useful for consensus purposes). FrozenPurpleCube, I will try to dig up some examples; any idea what you're looking for here? Carlossuarez46, your other suggestions, particularly numbers 11 and 12, seem to bear further discussion. I wonder, though - would some of them fall under a criterion already outlined in another policy or guideline? That'd be something to explore, but they're certainly valid concerns.

• To the group in general - It seems to me that there is little objection to WP:LISTCRUFT criteria 3, 4, and 6 aside from clarification of scope - particularly number 6. Any comments there?

Altogether, this is a VERY productive discussion. If we can continue to keep it civil and assume good faith from one another, I am confident we can arrive at a consensus that could allow us to clarify WP:LIST quite a bit.

And with that, I am off to bed! Thanks again for everyone's input so far! :-) Sidatio 03:26, 14 August 2007 (UTC)

I'm thinking of lists that were in a poor state, but were improved, or list ideas that were found to be overbroad. FrozenPurpleCube 03:51, 14 August 2007 (UTC)

I don't like WP:LISTCRUFT #3: WP:NOT#INFO is used like a machete against lots of lists, discriminating or not. #4 is a bad combo: unverifiable I agree with totally, but WP:NOTABILITY has no concept of what list articles are, or what they are for. #6 is okay, but "unmaintainable" could mean anything. The only two criteria I could agree with off the bat with are "Unverifiable" or "Too broad in scope", which are sensible criteria frequently used at WP:CSD. Oh, wait, I'd agree with "violates NPOV" (attack lists, etc.) as well.

I like this part of WP:LISTCRUFT: In general, a "list of X" should only be created if X itself is a legitimate encyclopedic topic that already has its own article. In particular, I see no harm befalling Wikipedia from having "X in popular culture" articles, as long as they're verifiable (but not necessarily verified) and not so broad in scope as to be more than, say, 200 entries long.--Father Goose 06:15, 14 August 2007 (UTC)

[edit] Proposal for a very short guideline/policy statement

I like the enthusiasm everyone's had on this subject. I believe that the most important words said on lists come from Listcruft:

In general, a "list of X" should only be created if X itself is a legitimate encyclopedic topic that already has its own article. The list should originate as a section within that article, and should not be broken out into a separate article until it becomes so long as to be disproportionate to the rest of the article.

I'd use this (to me totally sensible) guideline to form a policy based on three points:

1. Each list should begin with a link to the (single) article which it supports. Usually a list should begin as a section in that article.
2. Lists should be long enough that their inclusion in the supporting article is not feasible. Usually that would mean about ten items or so, but it could be less for a mature article at or beyond the length guidelines.
3. Lists should have under one hundred entries (or, better, lines of text). Longer lists tend to become indiscriminate collections of information that are hard for readers to digest at a sitting. Longer lists can be broken up to sub-lists by category provided the sub-lists can also be linked to an article where they could be included. (E.g., "List of U.S. Representatives from Ohio" could have a sub-page "List of U.S. Representatives from Ohio, 1800-1900" if there were a page "History of Ohio in the Nineteenth Century" to which it could be linked)

What the first and third categories ensure is that Wikipedia doesn't forget its mission of being a (prose) encyclopedia. If we find too often that we have lists which have no articles to refer to, it should indicate that we're spending too much time on list making and not on article writing. "Women and Women's Issues in Iran" is an article I think would be a great addition to this encyclopedia, and it would make the "List of Iranian Women" have a solid foundation for showing its relevance to the project. -- Myke Cuthbert (talk) 03:53, 14 August 2007 (UTC)

Mscuthbert makes a mostly sensible suggestion here. Some lists, however, are necessarily a bit longer than 100 items, such as List of female composers, while providing an excellent (and constantly improved) reference that exists nowhere else. Badagnani 03:57, 14 August 2007 (UTC)

The first part of the proposal is the one I care most about, and I agree that List of female composers is a great list (but I think proposal 1 highlights our need to have an article on Women in Music or on Women Composers). But once it becomes extremely long, it's hard for anyone to have complete expertise over. At this point, to me, the list has enough depth that I think we should have a separate article on Women in Music in the 20th Century. I think it is a case of the list getting larger than its prose context. (P.S.: Is Vanessa-Mae a composer?) -- Myke Cuthbert (talk) 17:29, 15 August 2007 (UTC)

She is indeed. I agree that it's a long list, but what can we do with it? Being a female composer is clearly a notable criteria for a variety of reasons (the main one being that it's very common to point out "hey, x composed music, and was a woman!" in various places), and I don't think a simple "women in music" would cover it well -- as that would theoretically include singers, where being female doesn't have nearly the same level of notability over being male...of course this could be noted in the article. There's various ways of going about it, but it should be in addition to a well made list, not instead of it, IMO. ♫ Melodia Chaconne ♫ 19:17, 15 August 2007 (UTC)

If the current proposal passes and is posted, it would cause List of female composers to be reformatted as an index list (or "list of lists), much like Lists of philosophers. It would have the end effect of making the list more searchable, instead of having to browse through a ton of entries to find what one is looking for. List of female composers already seems to be subdivided, so it should be relatively to sub-list based on time period. What do you think? Sidatio 19:24, 15 August 2007 (UTC)

Nothing against that in my mind, at least, as long as it doesn't set itself up for deletionists...♫ Melodia Chaconne ♫ 19:43, 15 August 2007 (UTC)

If anything in my policy proposal makes it easier for people to delete information on the contributions of women composers, then I want it withdrawn instantly. I think though that long lists are more open to deletionist attacks, because once a list gets above a hundred or two hundred entries, it begins to at least have the appearance of being open-ended and unmaintainable. While a shorter list clearly supported by (and branched off from) a notable article seems more notable in itself. -- Myke Cuthbert (talk) 02:24, 16 August 2007 (UTC)

Since there would be something official to conform to and cite, I wouldn't think there'd be much cause for worry in that department - so long as the list conforms to policy and guidelines, what's to delete? However, this is the short-term fix. The long-term fix is to eventually replace most of the larger lists like this with categorization once categorized lists become as visible and readily searchable as the current, manually-maintained lists such as List of female composers. Since this will take quite a while to implement, the conventional wisdom is to adapt policy to allow these lists to survive until then. (It's better explained at User:Sidatio/Conversations/On list guidelines#Categories need to be a tool, not a source of navigation frustration.) Sidatio 19:53, 15 August 2007 (UTC)

I was asked to comment and I find the above discussion quite interesting if a bit hard to follow. I originally closed the Iranian women AfD and I'd first like to repeat here what I said there. A guideline could be useful if only to avoid going back and forth between CfDs where people argue for listifying and list AfDs where people argue for categorizing. In many ways, I think the case of Iranian women is a great case study. There are those who argue that this list quite simply contains information that a category can't. That's certainly true. Others argue that this is pretty close to violating WP:NOT which I think is also true. However, in that particular case and in fact in many others, there is a possible compromise to reconcile these two positions and that is creating topic-specific sublists which also provides sufficient context. The problem is that doing this properly requires a lot of work and, as is often the case, there are just not enough people taking the time to do the work. Another recent example is Wikipedia:Articles for deletion/List of hotels in Hong Kong: I suggested then creating a page similar to Hotels in London. Again this requires more work but is a solution acceptable to most. If we are to build a consensus, we can't aim for some sort of rulebook clearly defining a set of absolute criteria. We need to find a way to explain when and how lists of more dubious value can be transformed into nice encyclopedic articles. The key, in my mind, is the following test: if you find it hard to write two or three lead paragraphs explaining the context around the list then chances are the list should go. And if you are able to write these two or three lead paragraphs well then write them! Pascal.Tesson 04:39, 14 August 2007 (UTC)

Something similar to this--the subdivision into smaller lists--is happening to the deleted List of Indian women. Several of the categories I created from that list have not survived current policy, and they are being listified. The new lists are, obviously, smaller in scope: List of Iranian women athletes, for instance. --Moonriddengirl 12:21, 14 August 2007 (UTC)

I'm responding to a discussion canvass as well. Unfortunately I have been much too busy with real life demands so I will basically second what Pascal has said -- I see too many lists being kicked back and forth between CFD and AFD, and a lot of gray area lists receiving inconsistent closures due to the lack of clear guidelines. I am also concerned as Sidatio has been that there are deletion rationales being given without guidance or basis in policy or guidelines, amounting to WP:IDONTLIKELISTS. Since maintainability and usefulness are often brought up in list deletion discussions, these should be incorporated into any guideline after consensus (even though WP:USEFUL is considered harmful, more or less, it seems to apply more to lists than any other type of article). That's about all I can contribute for now. --Dhartung | Talk 05:19, 14 August 2007 (UTC)

The gist of Pascal.Tesson's proposal is that all stand-alone lists should be deleted unless someone writes 2-3 paragraphs to turn them into articles instead of lists, is that correct? Kappa 07:34, 14 August 2007 (UTC)

Why on earth should list of glaciers start with an introduction to glaciers, duplicating glacier? If people want that, they click. The functionality of lists is to list links. Added value can always be sought. But the criterion for inclusion in WP should be the contribution of the plain listing to WP. Charles Matthews 09:25, 14 August 2007 (UTC)

Oh come on Kappa, you know that's not what I'm saying. A list of glaciers does not need to start with 2-3 paragraphs duplicating the glacier one. However in many cases where there is no obvious parent article, this is (as I said earlier) a good test: is there a way to put this list in proper context. If not, this suggests a list in which items have too tenuous a relation to appear in a meaningful list. Pascal.Tesson 02:41, 15 August 2007 (UTC)

New comment In response to Myke Cuthbert's ideas--

# Each list should begin with a link to the (single) article which it supports. Usually a list should begin as a section in that article. I think this is an excellent suggestion.

# Lists should be long enough that their inclusion in the supporting article is not feasible. Usually that would mean about ten items or so, but it could be less for a mature article at or beyond the length guidelines. I'm on board with this one, too. :)

# Lists should have under one hundred entries (or, better, lines of text). Longer lists tend to become indiscriminate collections of information that are hard for readers to digest at a sitting. Longer lists can be broken up to sub-lists by category provided the sub-lists can also be linked to an article where they could be included. (E.g., "List of U.S. Representatives from Ohio" could have a sub-page "List of U.S. Representatives from Ohio, 1800-1900" if there were a page "History of Ohio in the Nineteenth Century" to which it could be linked) I'm not sure about this one. "List of U.S. Representatives from Ohio" seems to me to be a manageable list regardless of size in part because criteria are obviously clearly defined. I worry that arbitrarily dividing lists can lead to article explosion. But, on the other hand, dividing up the List of Indian Women seems an excellent solution. I think it might be good to qualify this approach with a note on the breadth of the topic. --Moonriddengirl 12:21, 14 August 2007 (UTC)

[edit] On the contrary

If I may chime in with a simple proposal, I think that a list does need an "introduction to the topic", but it is fine if that introduction is in another article (no need to copy/paste glacier to the top of list of glaciers, just link it). I believe the main distinction between lists and categories is the added information. If a list is nothing else, and cannot be expanded to more, than an alphabetical grouping of links, then it should be a category instead. If, on the other hand, there is more useful information available (e.g. a "list of Some Film Award winners" could include the years in which the people won, as well the name and genre of the film in question) it should be a list rather than a category.

It makes sense to keep short lists in the "introductory" article, and move out longer lists. It goes without saying that a list of e.g. people should include only those people that have, or could plausibly have, an article on Wikipedia (i.e. no fair adding your neighbor to the list of soccer players just because that's his hobby). >Radiant< 11:38, 14 August 2007 (UTC)

Comment: I think you make a good point, although I doubt there's likely to be too many lists that you couldn't expand. Take List of Beatles songs, for instance--straightforward listing. Look what's been done with it at List of Beatles songs written by Lennon/McCartney. :O (Major effort there.) I think it would be a good addition to the criteria, but probably as one in a group and specifically not in itself a "saving point." I heartily agree with your opinion on notability of entrants and am inclined to think we should default to bluelink only, unless an entry is sourced on the list. Verifiability is a core policy at Wikipedia, and an isolated name is not verifiable per policy. --Moonriddengirl 12:35, 14 August 2007 (UTC)

I can think of several lists that don't have much of a meaningful expansion, such as "list of <nationality> <profession>". >Radiant< 12:40, 14 August 2007 (UTC)

You might be surprised. :) If you look at List of Iranian women, it's subdivided by profession, and they've got additional information on every one. --Moonriddengirl 12:50, 14 August 2007 (UTC)

"Woman" is not a profession, last time I checked :) >Radiant< 13:28, 14 August 2007 (UTC)

Um, no, but "subdivided by profession" does sort of cover that. :) Point being, there's plenty to say about nationality/profession. --Moonriddengirl 16:27, 14 August 2007 (UTC)

Bluelink does not meet verifiability, not even for lists. You might want to take a look at WP:ASR. Also, the reference section for List of Iranian women was blank before and after the meaningful expansion. Did Wikipedia gain or lose on that one and was the expansion really meaningful? -- Jreferee ^(Talk) 13:35, 14 August 2007 (UTC)

What would you propose, then? Clearly it's not practice on Wikipedia to reference items in lists. Take List of mountains as a purely random example (did y'all know we had a list of mountains? off the top of my head; not surprised we do)--not a single reference in sight. Bluelinking as a guideline may not meet verifiability per WP:ASR, but at least it offers some speedy way to verify. Current practice is entirely "take our word for it." :/ --Moonriddengirl 16:33, 14 August 2007 (UTC)

[edit] Day 2

Looks like we've got a fair amount of commentary out there, from a wide range of schools of thought. I've read through the comments, and to me, it seems like this is the direction the discussion is taking:

• Maintainability: With few exceptions, there seems to be a consensus that The list is unlimited and/or unmaintainable is a valid criterion if properly defined. However, there's not much of a consensus on what to do with a list that fails to measure up to this standard.

My take: Speaking as a new-ish Wikipedian, I'd have to say my chief complaint is how difficult it is to search categories. From what I've seen so far, I have to either type "Category:whatever" in the Search box, or go into a related article and hope it's in the category I want. Searching for the topic alone (say, "American people") brings up the huge honkin' list of lists that has to be maintained by hand. It seems to me that, if we had a better search function for categories, a great many lists simply wouldn't be needed, freeing up manpower to create more articles. (On redlinks in lists - On one hand, they may foster article growth. On the other, those redlinks require research to see if they're viable. When it comes to patrolling recent changes vs. patrolling new articles, the latter seems to be a LOT easier than the former to me. Can any CVU members in the audience give us some of their insight?) So, until that function is implemented, it can be argued that lists of most sizes serve a purpose as readily searchable categories.

Possible solution: Keep larger lists until a better way to search categories can be implemented, but make sure those lists are sub-listed - much like Lists of Americans. I think a good standard to aim for (at least with people, which seems to be the largest issue) is a minimum of three inclusion criteria, like "list of Armenian women nurses", in order to be a stand-alone list - that is, to avoid sub-listing like the above example. By the same token, perhaps a maximum inclusion criteria standard should be reviewed to avoid things like Wikipedia:Articles for deletion/List of songs by African, Asian, Caribbean and Latin American artists which reached number one on the Hot 100 (US).

QuestionWe don't know if the category system is ever going to be fixed or improved though, thats the main problem. If we don't get a new system for categories, what's our next step?

• Good question. That would be the next step in the process, after clarifying WP:LIST - we'd have to ask for the function to be implemented. In order to do that, though, we'd have to establish consensus that such a step would be welcomed - setting up and modifying search databases on a site like Wikipedia isn't something to be taken lightly, as it involved a lot of work. In the interim, our best bet would be to find a solution that could serve as a long-term fix if needed. The outlined possible solution for this issue could conceivably do just that. Sidatio 13:34, 14 August 2007 (UTC)

(rereads the possible solution) Three...yeah, I would say thats okay. That would actually work perfectly. For a maximum amount...I would say 6. Thats enough for some more specific stuff, but anything beyond that is too specific.Silver seren 13:37, 14 August 2007 (UTC)

In terms of CVU, yes your right. I almost never use recent changes anymore. Exclusively, I use the new pages or the new users to look for stuff. This can be seen by my user page. Though i'm not clear by what you want from this...Silver seren 13:13, 14 August 2007 (UTC)

• Viability: >Radiant< raises a good point in regard to list lengths. Let's take, for example, the List of Iranian women, which would be sub-listed if the first solution is implemented. That would leave three Iranian women athletes, which would make for one VERY short list.

My take: I don't think it's unreasonable to set a standard for the number of entries in a list. If there's, say, less than 15 entries that can go onto a list, it seems to be prudent to roll those entries into a parent article until such a time as there are enough entries to split off into a new list. In the example above, we could roll those athletes into Sport in Iran. This eliminates a list and bolsters an article at the same time.

Possible solution: Set a minimum entry standard for list creation and retention. A maximum entry standard could also be reviewed, but may be difficult to set and may already be addressed if the first solution is implemented in some form.

Agree Yes, that seems like the best way to deal with situations like that. The main problem would be that some will disagree, thats an unalterable fact. I suppose we'll just have to deal with them when the time comes.Silver seren 13:16, 14 August 2007 (UTC)

• Lack of clarity in present guidelines: The wide range of opinions on the topic seems to be clearly indicative of a need to further clarify WP:LIST. Further clarification is also necessary to avoid improper implementation and citation of essay guidelines like WP:LISTCRUFT.

My take: We're well on our way to laying the foundation for a better WP:LIST. However, we have to be careful to keep the future in mind - if searching Categories became easier, a great many "List of lists" would almost certainly be better maintained by category, especially since the "redlinking factor" wouldn't be as much an issue (if at all, since lists already tend to be developed before inclusion on a "list of lists). If a better search function were to be implemented, we would be wise to take that into account and establish consensus now to deal with this contingency. That way, a guideline on the topic would be readily available, and wouldn't catch the creators and maintainers of such "list of lists" off-guard.

Possible solution: Continue what we're doing here, with an eye on future as well as present issues.

Then we'd have "Night of the Living-Dead WIkipedians", huh? Silver seren 13:38, 14 August 2007 (UTC)

As always, I eagerly await your responses and interpretations (provided they're civil). Sidatio 13:08, 14 August 2007 (UTC)

Some thoughts (Moving my comment down here, as it does have a couple new points)I found this page because of the PROD at List of female composers mentioned above (which has since been removed by Radient). I think it's a perfect example of the problem with these proposals.

• Question - It seems a strange coincidence that the PROD of List of female composers went up just hours after I mentioned this article here. I don't think I'm going to be able to discuss these issues any longer here, with editors such as these. Thanks for the invitation, but...sorry, goodbye. Badagnani 16:41, 14 August 2007 (UTC)

• Comments. Respectfully, I think this is one of the main reasons this conversation needs to happen. If there were a clearer policy, PRODs and AfDs might not be as contentious as they are. --Moonriddengirl 16:49, 14 August 2007 (UTC)

I think the bigger issue, perhaps, is this -- SHOULD WP be used for such? The way WP works, it can be an almost godsend to finding a...how to put it...easy to find reference that doesn't exist anywhere else (or does, but can't be updated because it's only in hardcopy). A good example is Symphony No. 7 (Sibelius) discography, which is basically a list. From what I've seen, most people here would want it deleted...but to me, a discography of a classical work is quite similar to one of an artist, which is VERY common on musician articles (and before you say something, consider that a complete discography of some classical composers would number well over 5,000, even for CD alone for a few composers).

I think part of the issue is that people WANT this collection of info (I'm avoiding the use of the vague term 'indiscriminate' here). Since WP is so large, it's got one a way of being visible to let people update, as well as an easy way to find the info to refer back to it. This is why there's SO MUCH friction, especially with fictional articles, even discounting obvious spam.

And the redlink issue is a BIG reason why lists are so useful. They let people see the holes. Maybe require sources for justification of them? Let's say someone wanted to add Louísa Leonarda to the female composer list (a random redlink from it I picked out). They could need to find a source that at least said she was a female composer. Thoughts? ♫ Melodia Chaconne ♫ 14:02, 14 August 2007 (UTC)

Comment: Discographies of classical scores are unique cases, as far as discographies in general are concerned. Quite a few of them are so large, that they can be spun off into their own separate lists, as evidenced with the Symphony No. 7 (Sibelius) discography. To me, this is another example of an issue that can be clarified by adding minimum entry standards. If the discography's contents fall under that standard, include it in the main list. If it's over, split it off, link it to the main article like so:

Main article: Symphony No. 7 (Sibelius)

and treat it as any other list.

As to the redlink issue - there is a good argument for its usefulness as a way to call attention to articles that need to be added. However, there's also a good argument that such a feature might not be worth the efforts it would take to patrol new additions to make sure they're notable. It's an issue that needs to be resolved before settling the "categories vs. lists" debate, certainly. We have the input of one CVU; I feel more input from counter-vandalism specialists and new page patrollers is the key to settling that portion of the debate. Sidatio 14:31, 14 August 2007 (UTC)

No, you don't need to "patrol" redlinks in that way, any more than on the rest of the site. Any time that anyone wikifies an article, there may be a redlink created that we don't need: i.e. creating a redlink is an assertion that a topic is worthy of inclusion, and that may be wrong. But we don't worry about that, mostly. If an article is created, and we do not need it, then it gets deleted and all links to it should be removed. But there is nothing particularly special about the case of lists, that I see. It is helpful if some reference is added along with a redlink, in non-obvious cases, but why not treat these to a {{fact}} template in a normal way? Charles Matthews 07:01, 15 August 2007 (UTC)

Addendum: List of mountains is another great example of a list that is best maintained by a category; however, because categories are so difficult to search, it's going to be hard to find if dismantled. Again, the short-term answer seems to be sub-listing, and the long-term answer to seek a revamp of category searching. This might be a key part of developed consensus when we're finished here; can I get some more input on this and the other proposed solutions? If these pass, we may be able to wrap this up in a day or so, which would be one helluva feat considering the subject matter! Sidatio 16:54, 14 August 2007 (UTC)

[edit] Categories need to be a tool, not a source of navigation frustration

I would suggest the following principles:

1. We must dispel the notion that lists in and of themselves are not useful information, e.g., should exist only to link to articles--although that is not a bad criteria for practical--as opposed to informational--usefulness. A list which does not contain any items which are noteable enough to have their own articles is probably not encyclopedic material.
2. We must dispel the notion that size matters. Cutoffs in either direction, being artificial boundaries, are therefore also arbitrary.
3. We must work toward an "inventory" standard: guidelines, manual of style, helpful bits of automation/organization--not a list (!) of rules under which information is to be deleted.

A particularly thorny issue is that categories are unwieldly, frustrating to find, impossible to search, and extremely unfriendly to navigation once you get there. I've long since given up trying to use them for anything--perhaps they've improved, but I see them no more than a tag on an article. So, the first proposal is:

• Permit categories to be navigated as simple lists, that would go a long way toward lists aimed in that direction. Finding a category, then having to click on a letter of the alphabet to navigate to two items in a list of 20, 50, or even 100 is incredibly frustrating.
• If that is done ("flat view"), then categories will become much more useful. Many currently manual lists could be generated from the Category tag, again, using my example, Category:Latvians then generates the view "List of Latvians" from all articles tagged with the category (and perhaps a bit of extra information from the template in each article, say a ubiquitous "80 character limit" descriptor).
• Categories now become a constructive means for encouraging adding to/creating "lists" through the creation of articles.

If a person is happy to navigate a list of 500 or 1,000 items or more in a "flat" view, I see no reason to not allow that, nor any reason to suggest that a list that size is not useful. (For example, copy into a spreadsheet or personal database.) —  Pēters J. Vecrumba 14:30, 14 August 2007 (UTC)

Comment: In order:

1:I concur. Lists on the whole aren't necessarily bad things. There are, however, bad lists, as Carlossuarez46 outlined yesterday and Pēters J. Vecrumba concurred today.

2:At some point, maintainability has to be addressed. Imagine, for a moment, the manpower required to maintain a list of notable Asians. They're about 1/3 of the entire world's population. It's a Doomsday scenario, yes, but even lists of people by country can become overwhelming if not sub-listed. Until and unless categorization is improved, sub-listing seems to be the most efficient way of maintaining large lists like these. I also fail to see how a definite number is "arbitrary".

3:That is, in fact, what we're doing here. There are indeed "deletionists" taking part in the conversation, but I don't see much serious call for wholesale removal of lists on the whole.

I wholeheartedly agree with making categories easier to search. Even if it were just another box under the main Search box labeled "Search categories", that'd be a ton and a half better than what we presently have, and would greatly enable the categorization process to take over the maintenance of quite a few lists. It's definitely a future issue that we can help address now by requesting it. Let's get a little more input on how we'd like to see that function implemented, then go about requesting that it be added. Meanwhile, we should also keep our eye on the present issue of better defining current policy, which seems to be taking shape nicely. Sidatio 14:58, 14 August 2007 (UTC)

Ahhh... "List of (noteable) Asians"...

    Well, it seems something more granular (Japanese, Chinese, Indonesian, Thai,..) would work. But you do bring up the issue. Let's say we do get categories to work constructively, where they give us informative lists to easilty navigate. Those articles which all carry a tag for "Category X" -> give us "List of X".

    Now enter the doomsday scenario. While a list is going to be as long as it needs to be, there's nothing to prevent following some usability conventions where categories are concerned. So, for example, perhaps someday "List of Latvians" (generated from Category:Latvians) comes up with: "Category exceeds 1,000 items, only first 1,000 returned." Well, there's the incentive to keep things smaller. But partial data only upsets people. And, again, the list could validly be 1,000 or 10,000.

    There are two problems around sub-categorization:

1. what's it going to be? based on alphabet (uggh!)? based on some attribute--and if so, what?
2. touching all those 1,000+ (using an arbitrary cutoff) or so articles to enter the new subcategory data

    After spending some time ruminating over this, considering how to add "needs subcategorization" tags, how to get consensus on what subcategories should be, I realized that the answer to all the problems around creating subcategories was exquisitely simple: don't create any.

    Instead of subcategories, create subsets by other existing categories (rare is the article that once categorized appears in one only). For example, if List of Latvians is over 1,000 (and this could even be user selectable), instead, present a list of all other Categories Latvians appear in. So, for example, one could now choose to browse/list: Latvians + who were Victims of Soviet repressions + and who were not... You would then have the choice to view the resulting flat list or subset further based on mutually shared categories. (This, in fact, would be the model for browsing categories: give me a flat list or give me the option to subset by some other category; iterate until arriving at the desired data.)

    This would also make useful categories which we would never dream of creating otherwise because, by themselves, they would be their own a doomsday scenario: imagine "List of Doctors", "List of Politicians", etc. But now, Category:"List of Indian women" crossed with Category:"Doctor" (previously unthinkable!) would surely produce something quite useful and quite finite. More is less. —  Pēters J. Vecrumba 16:55, 14 August 2007 (UTC)

Comment: I like your ideas here on revamping categories, Pēters J. Vecrumba. Does anyone have any experience with databases and search functions? It'd be a huge plus if we could go to bat with some kind of understanding on how to implement these ideas.

Cross-referencing looks like a great idea. As I understand it, you're thinking of something like your standard Google search - typing in "Indian women" gets a ton of hits, but typing in "Indian women doctor" would significantly filter that out. However, something along those lines seems like it would take quite a while to code, and might not be compatible with the current search setup. How can we implement an easier category search that works with the current search system? If we can figure that out, it'll probably be easier to ask for, since it would be easier to integrate.

Also, I see what you're getting at with the "artificial limits" - you weren't referring to the idea for the lists, but rather the limitation of 1,000 entries returned by the current category search setup, am I correct? I think you're on to something there, but I also think this is definitely something that will be done well into the future. Good points.

I wasn't thinking of a multi-element database query, exactly. That presupposes we know what all the categories are. We'd still do an initial search somewhere to find our "top" category list (for example, List of Indian women generated by Category:Indian women).

    Clicking on the top level category list List of Indian women would come up with the choices of flat list or...

    in (perhaps choice of alphabetical or "most recently added") familiar chunks of "100 (20 | 50 | 100 | 250 | 500)", further categories that could be clicked on, for example, Women in war. That would bring up a fresh list of only articles which match both categories. A nested context would be kept allowing you to back up to any point in the chain (typical hierarchical navigation).

    "Crossing" only one category at a time with the rest found so far would reduce database overhead per query as well as (more importantly) intersperse human time with CPU time. Also, you often don't know what you're looking for until you see what's available. :-)  —  Pēters J. Vecrumba 18:11, 14 August 2007 (UTC)

P.S. Do we have stats on how many articles are in each of the top ten categories (whatever those are?)

Response: All I could find was the Wikipedia statistics section. Categories seem to be a truly neglected aspect of Wikipedia! Sidatio 18:46, 14 August 2007 (UTC)

I agree that the ultimate way to resolve the issue of having people create so many nearly useless lists is to greatly improve how categories work. Being able to search within categories and seeing a whole category either as a single list or one divided by alphabetical order would go a long way. --mav 01:24, 15 August 2007 (UTC)

[edit] Looking good!

There's enough here to begin working on a preliminary version of the new guideline additions. I'll try to get a rough draft of the additions up sometime late tonight. In the interim, if we could get one or two more comments on the proposed solutions (or any other proposals) on the short-term goal of clarifying WP:LIST, that'd put us over the top. It'll also give us better boundaries in the interim while we work on the long-term goal of better development and implementation of categorization and its search functions to eventually replace many of these lists.

This is some pretty heavy work we're undertaking here - I'd like to thank you all again for rising to the challenge. Sidatio 18:46, 14 August 2007 (UTC)

[edit] Maybe I don't get it...

And then again, maybe I do. I appreciate the invite to participate, but I might be guessing wrong here. From what I can figure out, you are endorsing the implementation by sharing the points contained in WP:LISTCRUFT in order (1) To show that the overuse of two syllable word "listcruft" is generally inconsistent with the definition of what listcruft actually is; or (2) to get people to question and debate the principles (clearly, some of the responders didn't recognize this as what's already "on the books", and thought it was your own suggestion); or (3) seeking to educate those of us who have been turned off from reading the principle by those zealots who shout "listcruft, listcruft", or who display WP:FIVE on a sheet at the stadium, next to the banner than says John 3:16.

My two cents worth...

1. 1 Wikipedia obviously does not prohibit lists. If that were the intent to do so, there would not be a statement about "listcruft", which distinguishes "bad" lists from "good" lists. As I have stated many times before, lists are not unencyclopedic. Reference books have appendices, libraries have card catalogs, encylopedias have tables. Lists are one of the most efficient ways of organizing information. Lists didn't start with Irving Wallace. God gave a list to Moses, Hammurabi gave a list to the Babylonians, Luther nailed a list to the door, etc.

1. 2 Yes, the guidelines to listcruft are subjective. AfD is where people venture their opinions about what Wikipedia should be. There would be no voting or discussion if it was left it simply to administrators to decide something runs afoul of Wikipedia. Hence, it is perfectly alright to venture the opinion that a list was (or was not) "created just for the sake of having such a list", or "of interest to a very limited number of people", or "cannot be expanded beyond a handful of terms", or "unlimited and/or unmaintainable". These identify what an article can be. I have no problem with any of these. I agree that we should all be more ready to defend our opinions that a list "was not" defective.

1. 3 Then, there are those criteria that identify what an article is right now. I think that these are misunderstood even by those persons who parrot them.

(a) Indiscriminate. My opinion is that most people who toss out the opinion that something is "an indiscriminate collection of information" don't understand what that means. Indiscriminate can mean random or chaotic, but the general idea is that the creator of the list makes no attempt to be "discriminating". We tend to use that word to refer to bigotry, but discriminating means that one has a standard with which to judge whether to accept or reject something (in this case, the information on a list). A good list has boundaries and doesn't include everything. Our experience with IPC articles (in effect, lists of pop culture references) shows that. (b) The content is unverifiable. This is curable. This bone is seldom tossed, but it does not mean that the mode of organization (the list) has to have already been in a published source. It just means that you have to be able to show, if called to do so, why "George Lincoln Rockwell" belongs on a list of famous Nazis. (c) The underlying concept is non-notable. This is a tougher concept. We have a rough test for notability of individual items, Ghits (i.e., how many times the phrase turns up on Google). Notability of a concept is, ultimately, a matter of opinion. (d) No content beyond links to other articles. This is something that can be fixed. The worst lists are the "list of blue-links" that reveal nothing about why something is on the list. These are crap. However, navigational aids are different. Some people freak out because they think these are "lists of lists", a double heresy if you hate lists to begin with. I like to say that, like an index in a book, these are not meant to be read for pleasure. (e) Would be better implemented as a (self-maintaining) category. When we vote on whether to keep or drop a list, we generally aren't aware that a category already exists. Categories, however, can be found in the same way as lists... there is a link at the bottom of an article. This is often a valid objection, and it's a substitute for the aforementioned "no content" list. On the other hand, some lists that do have content can co-exist with catergories. (f) The list is unencyclopaedic, i.e. it would not be expected to be included in an encyclopaedia. There's a lot of stuff in Wikipedia that one would not expect in an encylopedia, so it's difficult to argue this one with a straight face. What unencyclopedic means, in this context, is "not of interest". Anyway, I think I understand what you're seeking comment on. I agree that the rules of WP:LISTCRUFT do not merit deletion of most lists, but I also believe that we should try to follow the views of the creators of Wikipedia in voting what to keep and what to delete. Mandsford 01:04, 15 August 2007 (UTC)

Response: Actually, it was closer to your second guess than anything, though number 3 is funny enough to try. ;-) But seriously...

The main goal here is to set clearer guidelines for WP:LIST so we can avoid some of the inconsistencies that have been evident in list AfDs. At the same time, clearer guidelines should help take the bite out of some of the more contentious debates over some of these lists, and might also help cut some that have gotten a little out of hand down to a manageable size. Carlossuarez46 pointed me to a page recently called Special:Longpages. For those who don't know what this is, it's a list of the longest pages on Wikipedia, and a LOT of them are lists. Clearer guidelines could help break these down to manageable sub-lists in the short term, while we try to figure out how to request and implement better search features for categories that can eliminate the need for many of these list - and with them, many of the arguments and a TON of man-hours that go into list maintenance.

I think we have enough here to write a preliminary amendment to WP:LIST. I'm going to get started on that and hopefully have it posted in an hour or so on another subpage here. That way, we can review and tweak it before we post it. Meanwhile, Pēters J. Vecrumba has got some great ideas on the long-term solution of increasing the usefulness and visibility of categories, and I think we should continue fleshing that out. I'll try to find some programmers to chime in on the subject; if anyone else knows any, bring 'em in, and if you know what the guy's talking about (my expertise with websites ends with Dreamweaver), then chime in! Sidatio 01:51, 15 August 2007 (UTC)

[edit] PRELIMINARY GUIDELINE IS UP FOR REVIEW

User:Sidatio/Proto WP:LIST - the proposed changes are in bold and are in sections 2.3 and 2.4. Feel free to review them, and please suggest better wording; writing policy and guideline documents is not a strength of mine. :-p Sidatio 02:16, 15 August 2007 (UTC)

You'd probably not guess it from my usually informal approach, but in my day job, I'm an editor. I'd like to take a crack at some of this, since you seem open to that. :) I'm going to make sections and sign each to eliminate potential confusion in response. --Moonriddengirl 14:09, 15 August 2007 (UTC)

• Reading over this, it sounds way more pompous than I'd intended. I don't mention my day-job in a "and therefore I See All" kind of way. I meant it in a "I'm hopelessly nit-picky" kind of way. It's an apology, not an assertion of authority. :) --Moonriddengirl 15:32, 15 August 2007 (UTC)

• Information

  I think this would benefit by the addition of concrete examples. I know the problem with using examples is that a good article can be quite a mess by the time a user stumbles upon the policy, but I believe it would assist greatly in comprehension. --Moonriddengirl 14:09, 15 August 2007 (UTC)

• • Works for me - I'll dig some up either on lunch or tonight. Sidatio 14:35, 15 August 2007 (UTC)

• Navigation

  As is, I find this somewhat confusing. I wonder if it would be easier to follow like so--

Lists can be used as a table of contents. Users browsing without specific research goals are likely to use the See also lists. Users with specific goals that are describable in one or two words are likely to use the search engine box. Lists are valuable for users who have a general idea of subject but no specific terminology. Such users tends to use the lists of related topics (also called list of links to related articles). --Moonriddengirl 14:09, 15 August 2007 (UTC)

• • Looks good to me; anyone else? Sidatio 14:35, 15 August 2007 (UTC)

• Development

  just for clarity, I would propose replacing a word--instead of "not the main space," I think "rather than the main space" is easier to follow. YMMV. --Moonriddengirl 14:09, 15 August 2007 (UTC)

• • Again, good. The flow of a document is always important, as it helps enhance readability. Sidatio 14:35, 15 August 2007 (UTC)

• I like the bolded section under the criteria for inclusions. It might meet some dispute (since the aforementioned List of mountains would be right out, for instance), but it offers a clear policy. Likewise, the bolded section under Lists Content allows good sense interpretation but a good guideline. --Moonriddengirl 14:09, 15 August 2007 (UTC)

• • Okay, I should point out that I'm generally not awake until, say, 1 pm. Having said that, I'm afraid I don't follow what you mean by "List of mountains would be right out". Are you referring to the fact that, if these changes are implemented, that article would have to undergo massive restructuring? Sidatio 14:35, 15 August 2007 (UTC)

• Yes. Sorry for my lack of clarity. I forget that as sprawling as these discussions become, one little point made 7,200 hours ago might be forgotten. :D I mean that the current List of mountains would violate this proposed policy. That's not a problem for me. I think Lists of mountains is subject to precisely the kind of sprawl we've been discussing. List of mountains in Japan would be far more manageable, imo. (And, lo, it already exists.) --Moonriddengirl 15:31, 15 August 2007 (UTC)

Comment: How the hell many lists exist on mountains?

List of mountains would indeed face restructuring under the proposed guideline, and most "lists of lists" are overextensive and would best be served by categorization. However, as discussed here, categorization is woefully lacking in visibility and ease-of-use searching. Those issues will have to be addressed before we go categorizing and deleting these index lists; otherwise, they're going to be a pain to find AND there would be strong opposition to their categorization and removal because of the lack of visibility. That's why it seems the "short-term/long-term" approach is the best fit for this situation. We can eliminate a LOT of sprawl in the short-term, while keeping the information as visible as possible until the long-term solution of bolstering categorization's abilities can be finalized, submitted, and actually implemented. After categorization is upgraded, these lists can be categorized and deleted since the system will be able to maintain AND properly display the material that used to be maintained by hand. Sidatio 16:33, 15 August 2007 (UTC)

I didn't actually mean to drag Lists of mountains into the conversation. It was List of mountains I meant to focus on and only as it relates to the proposed policy. Typo. :) --Moonriddengirl 16:57, 15 August 2007 (UTC)

[edit] Day 3

So far, so good. We've identified short and long-term solutions to the issues at hand, and it seems we have consensus. As outlined above, I've put up the proposed additions at User:Sidatio/Proto WP:LIST for our short-term answer, and discussion should continue on the long-term goal of improving the abilities and visibility of categorization. Pēters J. Vecrumba has some good ideas on how to go about that; I think the next step is to bring in a few people with programming and/or search database knowledge to hammer out a way to make those ideas reality. I'll send out some invites. If anyone else knows of anyone who can help out, drag 'em in here and let's get this thing done.

The prototype list will stay up through the weekend, to give everyone a chance to comment. If all goes well, we should have guidelines in place by Monday morning that could not only help streamline lists on the whole, but defuse some of these nastier AfD situations. Sidatio 13:05, 15 August 2007 (UTC)

[edit] Day 4

Pēters J. Vecrumba and Father Goose have raised a valid point about the proposed "3/2" inclusion criteria approach. Please read the discussion as moved from the prototype WP:LIST talk page and chime in with your take on it. Thanks! Sidatio 13:05, 16 August 2007 (UTC)

[edit] About "joins" of list inclusion criteria

I think bolding and stating as an absolute that a list needs "3 criteria" for inclusion (to avoid amorphous masses) is too strict. For example, "Noteable Humans" >> "Noteable Americans" >> "Noteable Latvians" >> "Noteable Livs" -- yet by the stated guideline, these all count as just one criterion toward three. Lists should be given a chance to develop from fewer inclusion criteria as appropriate. —  Pēters J. Vecrumba 01:50, 16 August 2007 (UTC)

Every guideline has its exceptions. Also, the bold print is just to show what changes are proposed and where it's proposed they be placed in the guideline - it won't actually be in bold print when finished.

The reason I suggested 3 criteria as a stand-alone for people-related lists (and people-related lists only) is to avoid things like "List of Chinese women". Now, if someone wanted to come along and create a list of notable women from the Vatican, well, that's another story and a case could be made for exception. But the 3 criteria rule would just apply to people-related lists. To impose it on, say, geography-related lists would definitely be too restrictive. Sidatio 02:16, 16 August 2007 (UTC)

Nah, that's no good -- the 3/2 criteria rule shouldn't be a rule at all, at most a suggestion. List of Nobel laureates is perfectly appropriate as-is, for instance, although it does have supplemental sub-lists. If the aim is to keep lists under X entries in most cases, then phrase the rule that way, with the "criteria" as suggestions on how to keep things reasonable.--Father Goose 02:46, 16 August 2007 (UTC)

I was under the impression guidelines were suggestions, not really "rules" in the classic sense. That was the approach I was taking, anyway. Like I said, there are always exceptions. Sidatio 03:34, 16 August 2007 (UTC)

They have a "this is how we do things" role, and can be referenced to back up an action, so they have quite a bit of binding power. You don't want to get them wrong. Treat what you're writing as though it were an iron-clad rule, and if it's wrong, fix it, or change it altogether. You've got to get just about everybody to agree with what you come up with. It can't be wrong; it can't be "approximate" -- if you really want everyone to go along with it. This is a far steeper task than just coming up with functional ideas -- and even that is very hard. If you want to become a shaper of policy, or even guidelines, be aware: you've stepped in it.--Father Goose 06:34, 16 August 2007 (UTC)

Clarification: by "stepped in it", I mean no criticism: I'm saying that you're wading into something that's going to get your shoes very dirty. This is the easy part: brainstorming. It gets much harder just around the corner.--Father Goose 15:39, 16 August 2007 (UTC)

Ah, gotcha. Yeah, I can imagine actually implementing it is going to be the proverbial "Round 2". Sidatio 15:41, 16 August 2007 (UTC)

While I think saying I've "stepped in it" is a bit strong, everyone's entitled to their own opinion. :-)

I don't think List of Nobel Laureates is a good example in this case, because it's already exclusive by nature; very rarely do you get more than 14 or 15 entries for said list in any given year. Also, since it's already got natural subdivides (by award category), it would take minimal effort to split the list in the event it got too big. I think Pēters J. Vecrumba's view of this particular issue makes more sense from a practical standpoint. By all means, a list should be given the chance to develop from less criteria, and exceptions can and should be made in certain cases. From a discussion and debate standpoint, it seems making an exception for a guideline by loosening its standard a notch would meet with far less opposition and animosity than trying to restrict that standard a notch. Hence, the "3/2" approach. For example, how many notable doctors from San Marino can there be? I don't think a list like that would be a burden at all, and I think a person would be hard-pressed to make a viable argument against such a list having only two inclusion criteria. However, a list of notable Chinese politicians could get pretty big pretty quick, and would probably end up being held to the 3 criteria standard.

As to writing an "iron-clad rule": The discussion is intended to be only about amending a guideline, which by definition is flexible and allows for exceptions. From WP:GUIDE:

“

A guideline is any page that is: (1) actionable (i.e. it recommends, or recommends against, an action to be taken by editors) and (2) authorized by consensus. Guidelines are not set in stone and should be treated with common sense and the occasional exception.

”

Having said that, I'm all for better re-wording of the proposed language additions to better allow for exceptions, if that's the issue we're having here. Let's move this section of the discussion to the main discussion page and see what the others think. Sidatio 13:05, 16 August 2007 (UTC)

Having read that, how about modifying it as follows (proposed addition in bold):

“

Further, there should be at least three criteria for inclusion on a stand-alone list of people, or two criteria for inclusion on non-people related lists. For example, List of Armenian women nurses, or List of mountains in Peru. If there's two or less criteria for people-related lists, or one criteria for non-people related lists (ex. List of Pakistani women or List of Chevrolets), they should exist only as index lists, where sub-lists related to that particular topic can be archived. A good example of this is Lists of Americans. Of course, exceptions to these requirements can and should be considered, by consensus, on a case-to-case basis.

”

Comments? Sidatio 13:05, 16 August 2007 (UTC)

• I'm a big one for examples. How about:

“

Further, there should generally be at least three criteria for inclusion on a stand-alone list of people, or two criteria for inclusion on non-people related lists (Examples: List of Armenian women nurses, List of mountains in Peru). In most cases where there are two or less criteria for people-related lists, or one criteria for non-people related lists (ex. List of Pakistani women or List of Chevrolets), they should exist only as index lists. Sub-lists related to that particular topic can be archived on these index lists, much like the format of Lists of Americans. There will be cases where this general requirement will not apply - for example, List of Nobel Laureates, since Nobel prizes are already reasonably inclusive. As with all guidelines, exceptions can and should be considered, by consensus, on a case-to-case basis.

”

--Moonriddengirl 13:29, 16 August 2007 (UTC)

I added the List of Nobel Laureates example, but now I worry if we're getting too wordy for a single paragraph... I made a few more edits and split it into two paragraphs to kinda trim it down a hair and hopefully make it more readable. Anyone care to comment? Sidatio 14:01, 16 August 2007 (UTC)

Scans well to me. :) I have been wondering if it would be a good idea to test drive out with some random wikilists, to see how the new rules would or would not apply to them. If you think that would be beneficial, I'd be happy to scout about and see what I come up with. --Moonriddengirl 14:04, 16 August 2007 (UTC)

I have no objection to a test drive - it'd be good to see how it applies in the "real world" before potentially rolling it out. Sidatio 14:28, 16 August 2007 (UTC)

Comment From Category:Lists, I selected a few random categories to examine. It seems that most lists already conform to a “3 criteria for people”/ “2 criteria for non-people” standard, though there are plenty of exceptions. Some of them are just improperly named, like List of stone, which I’ve proposed retitling on its discussion page. (If nobody objects, I’m going to be calling it “List of stones used in architectural design”, I think.) (Another example: the surprisingly well organized List of chairs, which by its name alone sounds horrifically broad.) If adopted, the proposed changes will probably not result in a plethora of discussions about existing lists, but may result in some--and some of it might be merited. For example, we have:

(People)

• List of male television actors
• List of male theater actors
• Dance personalia
• List of guitarists

(Non-People)

• List of museums
• List of dances
• List of bridges

I'm not suggesting by inclusion here that these lists should be deleted; I'm only offering them as relatively random examples of lists that could inspire debate under the proposed guidelines. I think the criteria are workable. I also think that, like any guidelines, there will be gray areas. Are "comic books" two criteria or is it one? Does List of comic books conform? Should List of gestures be renamed List of hand gestures? And, really, is there any reason that List of Nobel laureates shouldn't be separated by category? (Is it better to have 6 separate lists--List of Nobel laureates in physics, etc.--or one large one?)--Moonriddengirl 15:06, 16 August 2007 (UTC)

Response: As far as something like List of comic books goes, I think it satisfies the two-criteria requirement; it's a book, but it's also a specific kind of book. As to some of the others - Wow. The List of guitarists would definitely apply. Something like that really should be broken down by type of guitar played and music genre just for the sake of navigability alone (not to mention the terrible sprawl!). Same for the museums, dances, and bridges - break them all by nationality, and suddenly you have a much more manageable, searchable format. The dancers could be split into dance style and profession within the industry (List of ballet teachers would technically be three categories: dance, dance style, occupation, in my view). Now, it's good you found the actors' lists. List of male theater actors and List of male television actors are already defined by three criteria: Gender, profession, particular discipline within profession, but are already getting pretty big. Here, one could consider the proposed entry limits on 2.4 to open a discussion on whether manageability could be improved by adding another criterion, such as List of male television comedy actors.

All in all, it looks like the criteria are indeed workable, and flexible enough to inspire compromise instead of contentious debate. We'll leave this up over the weekend, put up an RfC tag (I've been informed you have to leave it in place for a while in order to have it listed), and see if any further issues arise. If not, I think we've accomplished our short-term goal. Sidatio 15:38, 16 August 2007 (UTC)

LOL! Duh. I'm thinking of whether "comic books" is one criterion or two and completely assuming that "theater actors" is one. Of course, it isn't. So much for those examples. :)--Moonriddengirl 16:27, 16 August 2007 (UTC)

One of the most important things to do is not say anything (in the guideline) that doesn't need to be said. It's vital to not close the door on good but non-standard list structures or criteria. The Nobel laureates list is just right as is, and is paired with several more-detailed breakout lists by-award. Both types of lists are exactly right for our coverage of the subject, and if the guideline lays out any rules that imply that the Nobel lists are misformed (or any other good lists), the guideline is wrong. This is what I mean by treating it like an iron-clad rule: by writing it in the form of a rule, guideline or no, someone will treat it like it's an iron-clad rule and the end result will be bad. So you've got to limit yourself to only trying to fix problems that are definite problems, not come up with something that fixes a lot of things but breaks perfectly-functional parts of Wikipedia in the process.

Keep brainstorming, and try to isolate the exact problems -- if there are any that are truly, unambigously, problems -- and find exact fixes that affect only those problems.--Father Goose 17:07, 16 August 2007 (UTC)

I think you're right to a point. I also think the proposal addresses many of those concerns:

• On its possible interpretation as an iron-clad rule: I think the most recent additions of the examples and the explicit language As with all guidelines, exceptions can and should be considered, by consensus, on a case-to-case basis should help greatly to avoid that kind of situation. Of course, there's always going to be the possibility of that kind of problem arising, just like with any guideline, but the consensus process by and large seems to handle these issues.

• On continued brainstorming: I absolutely agree. This proposal needs to be put to the test as much as possible before it's implemented, if it's implemented at all. We should definitely avoid breaking what's working - like the list of Nobel Laureates. However, we should definitely address what's NOT working, like list of mountains or list of guitarists. (The latter nearly broke my browser trying to load.)

The most important part is yet to come - seeing how it actually applies. We've started on that, though, and it's reasonable to assume that sufficient testing can be done by Sunday night. We should have a pretty good idea whether or not this will work by then. Sidatio 17:36, 16 August 2007 (UTC)

Well, I'd say the thing you're trying to fix, at least with this portion, is that lists not be too long. The most direct way to address that is to say "lists should not be too long" and define what too long is (200 entries-ish? 200KB?). Then suggest the criteria as ways to reduce the size of only those lists that are too long. Otherwise you've got a cart-before-horse problem. The best lists Wikipedia can offer, like the Nobel lists, will be perfectly focused to begin with, and will need no additional criteria. Some lists might be perfect at exactly 1724 entries, and have no sensible sub-categorization, at which point you should be recommending breakdown by alphabetical or other sorting order (e.g., List of United Kingdom locations: Be-Bn).--Father Goose 19:28, 16 August 2007 (UTC)

I think part of what this is attempting to address is not just the size of lists, but their breadth. This is already addressed in the current policy, but interpretation and application seem inconsistent. For example, List of Indian women was deleted following AfD; List of Iranian women was not. The arguments were the same for both articles, and it's pretty obvious that they were created with the same rationale. These are not the only lists I've seen up for AfD, and the debates on some of them have been heated. I think the whole community would benefit by a more fully defined policy on lists. --Moonriddengirl 19:43, 16 August 2007 (UTC)

Agreed, and stated better than I could. Size is a factor, but large lists aren't so much the trouble as lists with overly broad inclusion criteria are. These seem to cause the most problems among editors, largely because of the current murky policy and inconsistent AfD conclusions. Sidatio 19:55, 16 August 2007 (UTC)

Start from first principles then. "Articles like List of Iranian women should not be allowed because..." and I guess the answer is "overly broads" (er, broad). Nobel laureates is not overly broad -- so it isn't the number of criteria that distinguishes between "too broad" and not. You have to come up with something where people who don't know the ongoing issues at AfD read your proposal and go, "oh, yeah, that makes complete sense"; not "Three criteria? Who made that up?" The criteria idea probably has some validity, but unless you can offer guidance which makes people say "I understand that completely", no lists of criteria will ever get off the ground. Re-think the problem, explain why it's a problem, explain when it's a problem and when it isn't -- and why -- before throwing a fix at it that breaks other things.

You should know from AfDs that rules, including guidelines, get quoted extremely literally, so if you put any malformed ideas into "policy", they will cause big problems. First, do no harm.--Father Goose 20:26, 16 August 2007 (UTC)

All right - how would you word the guideline? So far, both the 2.3 and 2.4 proposals address list length and inclusion criteria. What else, in your opinion, needs to be addressed or needs to be made more clear? Sidatio 20:41, 16 August 2007 (UTC)

I suppose breadth is like obscenity: in the words of Justice Potter Stewart: "I know it when I see it." Problem is, it's very hard to make rulings based on a criteria like that, particularly when Justice Stewart is not around. :) What's the problem with an overly broad list? What makes List of Iranian women unmaintainable, while List of Iranian women in politics is not? I nominated List of Indian women for deletion because the criteria it established on the day I stumbled upon it was essentially "women who were born in India, born to Indian parent(s) or who moved to India." I pictured a cast of millions, and this list immediately seemed unmaintainable and unencyclopedic. The problem is that "unmaintainable" and "unencyclopedic" are likewise somewhat vague. I begin to see what motivated Justice Potter Stewart. --Moonriddengirl 21:27, 16 August 2007 (UTC)

Now you've got the problem. I don't think it's necessarily unsolvable, but it'll take more work before some really solid answers might emerge. I've been wrangling a similiarly-elusive issue over the past two months, in the form of Wikipedia:Relevance of content. I don't suppose I could lasso you into commenting on that proposal?--Father Goose 22:23, 16 August 2007 (UTC)

There's a lot of back-story to wade through there! I read over the policy and put my two cents on the talk page, primarily regarding readability. I suspect what I've suggested is precisely what you did have until somebody suggested what you've got now. :) Reaching consensus is a sticky wicket. --Moonriddengirl 00:33, 17 August 2007 (UTC)

Wait, I'm confused - I thought the proposed addition to section 2.3 of WP:LIST addressed breadth. Am I missing something here? Sidatio 00:38, 17 August 2007 (UTC)

Maybe. Maybe I am. What? :) Truly, I'm not sure what you're asking. As I understand it, we're talking about the exclusionary criteria for lists (2 for things; 3 for people) and questioning the use and function of that. As I define it, that controls breadth. Is that what you're talking about, or are you talking about something else? If it's that, I'm attempting to define to myself the problem--what makes "List of guitarists" inappropriate where "List of American rock guitarists" is not. Why is "List of Nobel Laureates" acceptable if "List of guitarists" is not? Is there a clear way to articulate that? Or, like Justice Stewart, are we stuck saying, "I know it when I see it"? --Moonriddengirl 00:55, 17 August 2007 (UTC)

Okay, we ARE on the same page, then. I was a little confused there, but I see we're still discussing the same thing. Here's the way I envision it:

• Comparing "List of guitarists" to "List of American rock guitarists" - I believe the 3-inclusion proposal helps here because it cuts down GREATLY on the scope of "List of guitarists". Naturally, there aren't going to be as many American rock guitarists as there are guitarists on the whole. Using "List of guitarists" as an index list (or a table of contents, as it were) would improve searchability notably by breaking down the guitarists into further categories, making individual articles on the list much easier to find. Likewise, it would make maintenance of all the lists involved easier for the same reason - if they're easier to navigate, it stands to reason that it's easier to find a list to maintain.

• On the List of Nobel Laureates: Given the fact that it's a rather rare award (only 14-15 maximum issued in any given year), it stands to reason that the award's rarity is an unspoken inclusion criteria. Now, if it were a list of both Laureates AND nominees, we might have a problem. However, that's not the case. As the years go by, this list will definitely increase in size, and splitting may need to be revisited. However, that's a long way off and we're still working towards a long-term solution that may or may not make splitting that particular list moot anyway.

• As far as the "Stewart Method" goes: I'm not even beginning to kid myself into thinking this proposal is going to be perfect right out of the gate. Not even the US Constitution was perfect from the start. Consensus is going to have to be applied to varying cases over time, to find exactly where the finer limits are. With the proposed changes, though, that fuzzy area we're seeing in regard to lists gets a HELL of a lot smaller. Well, to me, anyway.

That's how I see it. Sidatio 01:26, 17 August 2007 (UTC)

That's probably a wise way to see it. :) It's my nature to try to anticipate and prepare for every objection, and it's not always possible. I think the 3/2 criteria is a sound premise, and the text of your proposal takes into account case by case variance. And I do think you're right; fuzzy area shrinkage = good. --Moonriddengirl 01:34, 17 August 2007 (UTC)

I don't believe this "3/2 criteria" is a sound premise, because there are too many counterexamples, and even the examples given aren't very helpful. Easy counterexamples: The featured lists List of countries (one inclusion criterion for non-people) and List of French monarchs (two inclusion criteria for people). Whether or not guidelines can have exceptions, the exceptions should be rare and occasional, not featured! Furthermore French monarchs doesn't even have a separate article (it's a redirect to the list), so the requirement that there be a link to only a single article to support the existence of the list doesn't really work either. Another example: List of television stations in California is clearly a valid list, but there is no article corresponding to Television in California, nor should there need to be; the list is a valid geographical list of television stations in California.

And the examples given don't really reflect current practice, or even desirable practice. There is no List of Armenian women nurses, but we do have List of Armenians and List of nurses, both of which seem to be reasonably maintained. There is no reason to believe that a list of Armenian women nurses would be a better list than, e.g. a List of Armenian nurses. These guidelines would suggest that we split a perfectly valid list into possibly arbitrary subsets, just so we have the correct number of inclusion criteria. And while List of Pakistani women doesn't exist, I could envision it being a valid list to supplement the article Women in Pakistan. List of Chevrolets could certainly be created as an annotated and structured list of the vehicles in Category:Chevrolet vehicles. DHowell 06:16, 17 August 2007 (UTC)

That's an absolutely critical point: Wikipedia rules have to reflect current practice, or at least desirable practice. You can't legislate what people should do, or it won't be adopted; you have to lay out what people do do, or at least would do.

It's also occurring to me that just specifying a number of criteria doesn't help you much anyway: List of blue-eyed American Congressmen is, as they say on WP:CFD, an arbitrary intersection. I think CFD (and their bible, WP:OCAT) is a good starting point for researching some general rules; they handle things pretty sensibly there.--Father Goose 07:10, 17 August 2007 (UTC)

[edit] The problem is lack of clear inclusion criteria for stand-alone lists

The main problem I see is there is no clear guideline for when a list is acceptable, such as the guidelines for articles (notability), biographies, or disambiguation pages. The stand-alone lists guideline, listed as a style guideline rather than a content guideline, gives some guidance, but nothing that can clearly stand up to deletion arguments which cite WP:LISTCRUFT, WP:OLIST, WP:NOT#DIR, or WP:NOT#IINFO. But the essays, as often as they are cited in AfD, simply don't have consensus, and the policies are only consensus insofar as the specific things listed under the "not a directory" and "not an indiscriminate collection of information" subsections. For anything not within those specific things, the policy is vague and open to wide and varying interpretations. With such vauge guidelines and policies, it is no wonder that there is never any consistency in AfD with regard to lists.

What we need is a clear content guideline for stand-alone lists. This should be worded similar to the notability guideline, e.g.:

"A stand-alone list is presumed to be appropriate if:

• There are similar lists in reliable sources.
• It concerns a topic or concept that is notable.
• It has objective inclusion criteria.
• It is not a list of questions and answers, or long sprawling statistics.
• It is not so short that it is more appropriate as part of a larger article or list.
• It is maintainable; overly broad lists should be broken up into meaningful sublists, and the main list should contain a list of those sublists.
• It does not promote a particular point of view.
• It is not based on original research; all information in the list should be verifiable.
• The list is not simply an alphabetic or random list of internal links; lists should have annotation and structure to avoid duplicating categories."

Emphasis should also be made that if a list does not meet these criteria, it should be improved, not just deleted, unless the list's topic makes this impossible (e.g. it is a POV or non-notable topic, or there is no possible objective definition). This is just a rough proposal, I'm sure it can be worked on. DHowell 06:16, 17 August 2007 (UTC)

• "It is maintainable; overly broad long lists should be broken up into meaningful sublists." Kappa 10:58, 17 August 2007 (UTC)
• Or this: "Lists that are too general or too broad in scope have little value, unless they are split into categories. For example a list of brand names would be far too long to be of value. If you have an interest in listing brand names, try to limit the scope in some way (by product category, by country, by date, etc.). This is best done by sectioning the general page under categories. When entries in a category have grown enough to warrant a fresh list-article, they can be moved out to a new page, and be replaced by a See new list link. When all categories become links to lists, the page becomes a list repository or "List of lists" and the entries can be displayed as a bulleted list. For reference see List of people which is a broad list split into specific categorical lists." Kappa 11:07, 17 August 2007 (UTC)

"Brands" aren't a good example because there are already a lot of lists of brands around. "Chefs" would be a better one. Kappa 11:19, 17 August 2007 (UTC)

By the way, alphabetical lists don't duplicate categories if the category has already been split. Kappa 11:23, 17 August 2007 (UTC)

[edit] Day 5

It's nice to see some fresh commentary on this proposal, especially from two noted inclusionists. However, I feel it prudent to put this up in bold AND in quotes, just so everyone understands:

“

This is NOT a proposal to modify a rule. This is a proposal to modify a GUIDELINE. It should not, under ANY circumstances, be worded or structured to mandate to ANYONE what they must or must not do. It should be used as guidance ONLY. Consensus will determine the finer boundaries - this proposal's mission is to simply narrow what is currently a frustratingly vague scope in WP:LIST that is leading to marked inconsistencies during AfD debates concerning lists.

”

I emphasize this so strongly because of this comment by DHowell:

The stand-alone lists guideline, listed as a style guideline rather than a content guideline, gives some guidance, but nothing that can clearly stand up to deletion arguments which cite WP:LISTCRUFT, WP:OLIST, WP:NOT#DIR, or WP:NOT#IINFO.

Consensus will, and always should, dictate what will and won't be deleted. The goal here is to narrow the guidelines some, not absolutely, to avoid the craziness that has been list AfDs. (AfD in general is crazy, I know, but list discussions have been particularly erratic.) This is NOT going to turn into an attempt to influence a guideline to favor either deletionist or inclusionist viewpoints, nor is this going to turn into an attempt to make a guideline into a policy. You'll notice that effort in the wording of the proposed amendments - there's noting in them to state what MUST be done with lists that fail to meet those guidelines; it only offers suggestions. I firmly believe that neutrality MUST BE PRESERVED.

Having said that, let's address some of the points raised last night.

• DHowell makes a good point in regard to people-related lists with his citing of List of French monarchs. Looking over lists last night, it seems at first blush that just 2 criteria in most cases would be the best way to go for a stand-alone list. However, that's going to require defining what criteria is, and that could have its own issues. If we're going to go with 2 criteria all around, then gender shouldn't count as a criteria. That's just going to make the list too broad in scope, and we're going to have the same problems that happened with Wikipedia:Articles for deletion/List of Indian women and Wikipedia:Articles for deletion/List of Iranian women (2nd nomination), which is the aspect we're specifically trying to avoid. Over the weekend, if someone wants to help me take a comprehensive test of both the 3/2 and 2/2 criteria, get with me and we'll swap email, so we can better communicate our findings. We're starting to run out of room on this page.

As far as List of countries goes: It's a flawed example in the fact that countries are inherently inclusive. That is to say, they're really not making any more of them at a rate we'd have to concern ourselves with for decades. Now, List of mountains and List of lakes, while well-organized, are so large that breaking them into country or continent would be advisable from a navigation and practicality standpoint. Further, if an article's featured, then it's already been through the painstaking process that Featured Articles must take. It's already satisfied criteria that are above and beyond any guideline I can find, and are naturally not subject to lesser guidelines and policies.

Now, let's look at the proposed guidelines. I'm going to copy them down here, and give my take on them to the side:

• There are similar lists in reliable sources. - This looks like it would get shot down on sight by WP:OTHERSTUFFEXISTS.

• It concerns a topic or concept that is notable. - It should go without saying, really - ANY article is going to have to meet notability standards.

• It has objective inclusion criteria. - This is what we're trying to better define here. What is "objective"? Is it easier to make a list of what is and isn't objective, or just give a number as a guideline and let consensus determine the finer limits? Personally, I think the latter makes more sense; the former would more than likely end up being too restrictive for a guideline and might start crossing over into policy, which is impossible for us to decide. Policy, as I understand it, is defined by gaining either wide acceptance over a reasonable period of time by the community at large, or implemented by the Foundation.

• It is not a list of questions and answers, or long sprawling statistics. - Since these examples are already part of WP:NOT, it doesn't seem like there's much to gain from re-listing them in this proposal.

• It is not so short that it is more appropriate as part of a larger article or list. - Now, this definitely needs a number attached. The proposed addition to section 2.4 suggests 15 is the minumum.

• It is maintainable; overly broad lists should be broken up into meaningful sublists, and the main list should contain a list of those sublists. Again, this is going to need a number or it's going to get argued to death. The proposal for 2.3 suggests the 3/2 criteria, but we're also going to give the 2/2 variance its day in court to see which has less issues.

• It does not promote a particular point of view. - Already part of accepted policy.

• It is not based on original research; all information in the list should be verifiable. - See above.

• The list is not simply an alphabetic or random list of internal links; lists should have annotation and structure to avoid duplicating categories. - This contradicts Wikipedia:Categories, lists, and series boxes, which rightly claims that categories and lists can and do co-exist. The first portion is also better stated by WP:NOT#INFO.

So, much of what is suggested here is already part of official policy on article creation; restating it really won't serve to do much but add words to the guideline. The rest seems to have been covered sufficiently by the previous proposed additions to 2.3 and 2.4 of WP:LIST, as shown at User:Sidatio/Proto WP:LIST. However, the merits and problems of the "3/2" and "2/2" variants of 2.3 should definitely be scrutinized over the weekend. If anyone wants to help me undertake a comprehensive cross-section of existing lists to see how each variant performs (and I do mean comprehensive - I intend on looking at no less than 200 lists personally), then hit me up on my talk page and we'll exchange contact info so we don't clutter this page any further. I'll probably end up posting the finished research to my website for easier reference.

My response to your objections:

• There are similar lists in reliable sources. - This looks like it would get shot down on sight by WP:OTHERSTUFFEXISTS.

WP:OTHERSTUFFEXISTS is about other Wikipedia articles. This is about "other stuff" that exists outside Wikipedia, in reliable published sources. It is meant to get across that if reliable sources publish lists about certain things, then we should to, provided it meets other Wikipedia policies (e.g. WP:NPOV, WP:NOT).

• It has objective inclusion criteria. - This is what we're trying to better define here. What is "objective"? Is it easier to make a list of what is and isn't objective, or just give a number as a guideline and let consensus determine the finer limits? Personally, I think the latter makes more sense; the former would more than likely end up being too restrictive for a guideline and might start crossing over into policy, which is impossible for us to decide. Policy, as I understand it, is defined by gaining either wide acceptance over a reasonable period of time by the community at large, or implemented by the Foundation.

Objective does not necessarily mean "absolute" or "precise". One can have objective criteria which still allows for some subjective leeway, as I believe existing notability guidelines do. We do not need to make a list of what is objective or what is not, nor do I think we need to put arbitrary numeric limits on it; we can simply follow existing guidelines and policies. "Objective" in every other aspect of Wikipedia has generally been defined to mean "verifiable and attributable to reliable published sources", there is no reason for changing that with regard to lists.

• It is not so short that it is more appropriate as part of a larger article or list. - Now, this definitely needs a number attached. The proposed addition to section 2.4 suggests 15 is the minumum.

I disagree that this "needs a number attached". Attaching a number to guidelines without a clear consensus reason for doing so is arbitrary; and regardless of whether you say it's a guideline as exceptions there will be people who treat it as an absolute rule, and I envision people saying "This list has only 14 elements and therefore should be deleted if it is not merged with another article," or "This embedded list of 16 elements must be split out into a separate list article, and if that makes it non-notable then it should be deleted." Perhaps these are straw men, but I've seen similar things happen in AfD. "This number is too big or not big enough" are not usually acceptable in deletion debates, and guidelines attempting to set arbitrary numerical limits would seem to me to encourage such arguments.

• The list is not simply an alphabetic or random list of internal links; lists should have annotation and structure to avoid duplicating categories. - This contradicts Wikipedia:Categories, lists, and series boxes, which rightly claims that categories and lists can and do co-exist. The first portion is also better stated by WP:NOT#INFO.

I agree that categories and lists and do co-exist, however, I was simply trying to address the oft-repeated argument that a list should be deleted because it "duplicates a category". If that is an invalid argument, it ought to be added to WP:ATA. However, if "duplicating a category" is a legitimate problem, then fixing the list so it isn't duplicative is the answer, rather than deletion. Adding annotation and structuring the list in a way that categories cannot are ways of fixing such a list. Also Kappa brought up the excellent point that even alphabetical lists aren't duplicative if the category itself has been split. This does need to be worded, but the "delete because the list duplicates a category" or "delete because it would be better as a category" arguments need to be addressed somewhere.

Finally, to the argument that much of these criteria simply repeat other policies, I say that it is necessary, because it is not always clear how those policies apply specifically to lists. If it was clear, then there wouldn't be so much inconsistency in the way lists are handled in AfD. The guideline should repeat policies and guidelines which are applicable to lists and explain how they apply. For example, I've seen serious arguments that WP:N requires that the list itself be the subject of significant coverage in reliable sources in order for a stand-alone list article to exist; unless such an argument has consensus (which I don't believe it does), that ought to be addressed in the list guidelines. Hence the statement that the list should "concern a topic or concept that is notable", rather than requiring the list itself to be notable.

It is good that you are going to be studying a large amount of lists on Wikipedia; hopefully whatever improved guidelines we come up with will be generally in line with what consensus has already produced in terms of existing lists, and not end up disrupting the project by creating new debates about lists which wouldn't have existed without the new guidelines. I'm just afraid that that is exactly what arbitrary guidelines like "3/2" and "2/2" and "a minimum of 15 items" would do. The ideal guideline would reduce the amount of contentious debates in AfD, and possibly even convince the most hardcore deletionist that a certain list should stay, or the most hardcore inclusionist that a certain list does not belong. DHowell 21:27, 17 August 2007 (UTC)

[edit] Under new management

As noted on Sidatio's talk page, [they] have retired from Wikipedia to avoid Wikiaddiction. I'm all for having a guideline covering the ground everyone is working on here, so I hope the work continues. It would be good if one or more other users took the reins; I won't -- I've got my plate full, although I'll continue contributing.--Father Goose 02:49, 18 August 2007 (UTC)

I'm continuing with my project, test-driving the 3/2 guideline proposal. I've taken it through 10 out of the 50 pages, and will present on the experience when I'm done. --Moonriddengirl 12:12, 18 August 2007 (UTC)

But I should also note that while I plan to continue contributing, I don't think I'd qualify as "new management." I'm more worker-bee material on this one. :) --Moonriddengirl 15:02, 18 August 2007 (UTC)

[edit] test driving on lists of people

I volunteered to help Sidatio test-drive the proposal on the lists of people, looking for lists that did not meet the 3/2 criteria proposed, assessing solutions, considering the impact on the proposal in whole. I recorded my "journey" at User:Moonriddengirl/List test. It includes my final conclusions. In brief and with much less soapboxing these are:

• The 3/2 category guideline proposed presents considerable challenges for implementation and may be impracticable.
• Instead, it could be necessary to redefine the problem and find another way to articulate a solution.
• Any proposal should incorporate growth in consideration. Some lists begin small and general and should be narrowed as the list grows. This is mentioned in the current proposal, but may need specific application in content guidelines.

Please feel free to comment. I was asked to do 50 but am stopping at 20, primarily because I don't think that doing 30 more is going to make a difference in my report. Also, since Sidatio has retired, I do plan to undertake the lists of non-people as well. --Moonriddengirl 13:54, 18 August 2007 (UTC)

"Some lists begin small and general and should be narrowed as the list grows. If there are many items of a particular class, (eg. American), these can be split off into a new List of American X, while the remainder of the list can provide a home for non-American X's until their number grows. This process will of course be easier if the list is already grouped by nationality." Kappa 14:40, 18 August 2007 (UTC)

How would you apply that to List of guitarists? --Moonriddengirl 14:51, 18 August 2007 (UTC)

I would divide by genre; this looks pretty much like a List of rock/pop guitarists already. After that I'm not sure it would ever need to be split except alphabetically, because the category system is split by nationality and so it doesn't provide an alphabetical list (ie an index). If it was to be divided, I'd prefer to do it by year of birth than nationality but splitting off Americans, Brits etc would work. Kappa 15:22, 18 August 2007 (UTC)

I suppose so. (Alphabetical works for me. It's not pretty, but it's easily understood. :)) The existing policy covers starting broad and narrowing down, then. For me, I guess the challenge remains identifying what is too general and at what point generality becomes an issue. Having explored a whole bunch of lists now, I don't think a 3/2 category requirement for people is functional (List of kings of Rwanda), but I would dearly love to be able to look at a list and feel confident in knowing whether or not it is wiki-appropriate (List of people from California?). Based on the AfDs I have seen, I don't think I'm the only one uncertain about this. --Moonriddengirl 19:28, 18 August 2007 (UTC)

I've found alphabetical frustrating, as that basically assumes you know exactly what entry you are looking for (that is, split by alphabet only to make it more "navigable"). We need to think of people more like food, for example, would you want to navigate a list of salads by, for example, sub-list "Salads whose names start with the letter 'c'"? :-) No, you'd want to navigate by something that would narrow the list down in some meaningful fashion appropriate to the context of the list content. Context is king. —  Pēters J. Vecrumba 04:10, 21 August 2007 (UTC)

Alphabetical is a last resort. Subcategories/sublists, wherever possible, are preferable.--Father Goose 06:14, 21 August 2007 (UTC)

I agree with you both. However, subcategories are not always easily definable. --Moonriddengirl 11:25, 21 August 2007 (UTC)

But we should rise to the challenge instead of giving up and employing the "necessary evil" of alphabetical. When you're not quite sure what you're looking for or just browsing--which is exactly what encyclopedia readers do--scrolling through a list, no matter how large, is infinitely preferable to clicking on letters of the alphabet. If the reader knew what they were looking for, they wouldn't need to browse a list, would they? And lists as information lose their readability--and thus their usefulness--once you go alphabetical. —  Pēters J. Vecrumba 15:45, 21 August 2007 (UTC)

(resetting indent) Fair enough. How would you subdivide something like List of guitarists or List of saints? --Moonriddengirl 15:52, 21 August 2007 (UTC)

Index-type lists are definite exceptions. And having to split extremely long lists across several pages is a practicality -- obviously you can't don't want to put the phone book on a single page, online or on paper. If List of United Kingdom locations were put on a single page, it would take forever to load, and maybe even crash your browser. Alpha is good for huge indexes with no subcategories: all other types of lists should have as much internal structure/subdivision as possible. That's all we need to emphasize. We're probably all in violent agreement about this issue.--Father Goose 16:04, 21 August 2007 (UTC)

"That's all we need to emphasize." I agree with that. Alphabetical should not be the subdivision of first choice, but may be necessary when other options do not present. --Moonriddengirl 19:05, 21 August 2007 (UTC)

I examined your test drive (admittedly after the fact, well done! BTW) and will consider alternatives for List of saints. List of Anglicans and Episcopalians is another one that might be worth some attention, obviously smaller, but definitely suffering from zero-added-value "alphabeticalization." —  Pēters J. Vecrumba 22:47, 21 August 2007 (UTC)

(reindented) Quick question, I'm slowly getting to examining List of popes as a test case to try a couple of organizational approaches. Is there any way to generate a list of all categories each pope article appears in? Otherwise it will be quite a tedious exercise to create that... Thanks! —  Pēters J. Vecrumba 00:18, 7 September 2007 (UTC)

[edit] test driving on other lists

I am waiting to see if in the absence of Sidatio, who spearheaded this, there is enough interest to test drive the "2 criterion" proposal on non-people related lists. Obviously, it's a rather time-consuming task. Needlessly to say, I have no objection whatsoever to someone else taking it on. :) If no one else does and there is sufficient interest, I'll have a go at it eventually using a similar process as that recorded at User:Moonriddengirl/List test. --Moonriddengirl 15:52, 21 August 2007 (UTC)

I take your "list test" as proof that an X-criteria rule is not right for any type of list. Much better to try to find the answer organically. "Extra criteria" still has validity as a way to narrow a too-broad list, but only when the list really calls for it. And lists can start out with appropriate breadth and only expand over time into what should be multiple lists.--Father Goose 16:14, 21 August 2007 (UTC)

Boy, am I glad to hear you say that. :D --Moonriddengirl 16:56, 21 August 2007 (UTC)

I've been meaning to compliment you on doing that test, btw. I can see it was a lot of work, and analysis like that really helps to bring the problem, and possible solutions, into better focus.--Father Goose 18:56, 21 August 2007 (UTC)

Thank you. :) It was more of a headache than I expected, primarily because the further I went the more monumental the task of articulating this seemed to be. I'm starting a new section to discuss best ways of handling a new approach, sans criteria #. --Moonriddengirl 19:34, 21 August 2007 (UTC)

[edit] Evaluate WP:Not?

I begin to wonder if instead of building policy from ground up, existing policy should be simply altered with some simple additions. For instance, at the Wikipedia:Lists (stand-alone lists) page, down at Appropriate topics for lists, might it be helpful to simply insert something like As always when creating articles, we should be mindful of What Wikipedia is not? "What Wikipedia is not" could stand some clarification, especially WP:NOT#DIR and WP:NOT#INFO. Wikipedia:Lists (stand-alone lists) rather implies that breadth is not a major issue, since it lists as an example List of Elbonians. And, really, if the consensus on Wikipedia is that "List of Elbonians" and "list of Americans" and "list of guitarists" is okay, then maybe the problem isn't with Wikipedia:Lists (stand-alone lists) but with WP:NOT#DIR, since to my way of reading all three of those are "Lists or repositories of loosely associated topics." As I've become more involved with these rules, I begin to see why so many AfDs on lists become heated. I nominated List of Indian women for deletion (and it was deleted) based on my understanding of policy--it is a list of loosely associated topics. Most editors who commented agreed. OTOH, List of Iranian women survived, perhaps because more editors involved in that discussion were focusing on Wikipedia:Lists (stand-alone lists). I'm not here to promote rules that will send lists wholesale to the gallows (or, even, necessarily save them from it). I'm here to promote clarity on consensus to diminish the arguments. --Moonriddengirl 19:34, 21 August 2007 (UTC)

The entire "Not an indiscriminate collection of information" section is subject to overinterpretation. The other "not"s are more straightforward, but NOT#INFO simultaneously says too much and too little. It's a really coarse tool. How it should be changed, though -- I'm clueless.--Father Goose 03:19, 22 August 2007 (UTC)

NOT#INFO seems to be a big part of the problem in the "trivia" wars. I put my finger on what I perceive as the problem of NOT#DIR right here: "or persons (real or fictional)". In practice, WP is exactly that, ie List of Doctor Who supporting characters; List of guitarists; List of University of Oxford people. Each of those could be nominated for deletion in good faith by a reading of NOT#DIR (in spite of the rather vague "if their entries are famous because they are associated with or significantly contributed to the list topic"), while a reading of the "Lists of people" section at Wikipedia:Lists (stand-alone lists) might support them. --Moonriddengirl 14:19, 22 August 2007 (UTC)

Okay, that one needs fixing, then. We should try to establish some kind of consensus position on the issue here first, though.--Father Goose 17:52, 22 August 2007 (UTC)

[edit] Alternate proposal: clarification and alteration of up to three policies

Comment (with some retread for new arrivals) I think the two may be related. The problem as I see it is that we have three documents governing list contents: WP:Not, Wikipedia:Lists and Wikipedia:Lists (stand-alone lists). For the most part, the policies mesh (particularly Wikipedia:Lists and Wikipedia:Lists (stand-alone lists), though I think mention of the Wikipedia:Lists (stand-alone lists) policy could bear with emphasis at the Wikipedia:lists#Criteria_for_inclusion_in_lists), but there is contradiction implied in the handling of lists of people with WP:Not. (breaking for easier reading; I'm marking & signing each paragraph in this lengthy comment in case responses are inserted interstitially. One of my big problems with this discussion has been losing the flow; hope this helps! --Moonriddengirl 19:21, 22 August 2007 (UTC))

Comment part 2 Specifically, WP:NOT#DIR says WP is not for "Lists or repositories of loosely associated topics such as...persons (real or fictional)." It goes on to add, "Of course, there is nothing wrong with having lists if their entries are famous because they are associated with or significantly contributed to the list topic, for example Nixon's Enemies List." Applying that to List of Americans (which is a list of lists) will disclose a number of entries that seem to be in violation. Take List of people from Omaha, Nebraska--is Fred Astaire really famous because he's associated with Omaha? Has he really significantly contributed to Omaha? (Outside of maybe offering a tourist attraction, that is.) What about Marlon Brando? Is "Omahaian" the first thing people will think of when Brando's name is mentioned? Or is Omaha what it is today because of Brando's tireless contributions? The fact that these people are from Omaha seems patently to be a loose association. Wikipedia:Lists (stand-alone lists) by contrast would seem to make their inclusion okay...maybe. It does say that my neighbor can't be on a List of Christians because she's not notable for her Christianity, so it reasserts "notable for" as a criterion. But it goes on to add that "List of Elbonians would include persons who are famous in any category and who belong to Elbonia. The criteria for identifying as an Elbonian may not depend on the official citizenship laws of that country - the person could be related to the place by birth, domicile, parents, or by his personal admission, consider himself an Elbonian at heart." (breaking for easier reading; I'm marking & signing each paragraph in this lengthy comment in case responses are inserted interstitially. --Moonriddengirl 19:21, 22 August 2007 (UTC))

I'd prefer to see People from Omaha handled via its category, which is already far more complete. But for reasons mentioned by Vercrumba above, and some other reasons, Wikipedia's implementation of categories is poorer than lists in many ways. The three things I'd like to see added to the category code are:

• "Flat" browsing of subcategories
• Option to view the first ~100 characters of the article next to each item in the category (James Beard (May 5, 1903–January 21, 1985) was an American chef and food writer. He is recognized by...)
• Category-specific searching (American chefs AND (1800 births TO 1900 births))

If we had those three things, we could toss a huge number of completely unremarkable lists that are redundant with existing categories. Wikipedia's category implementation isn't remotely as useful as it ought to be.

Unfortunately, these are technical details. Can the servers handle it? Can we get the devs to do it?--Father Goose 21:27, 22 August 2007 (UTC)

Comment part 3 Why? If I rise to prominence as an actress (fat chance) and declare that I consider myself an Elbonian at heart, does that mean that I'm famous because of my association with Elbonia? Or that I've contributed significantly to Elbonia? If "persons by nationality" is an exception to the general rule of "notable for", then I think it should be explicitly stated as an exception. If not, then I think there needs to be clarification as to what it means to be "famous because they are associated with or significantly contributed to the list topic." I don't doubt that some of the cast of millions at List of University of Oxford people are famous because of their association and some of them have contributed significantly to Oxford University. Many of them are listed because they happened to attend Oxford, which seems to me like a loose association. (breaking for easier reading; I'm marking & signing each paragraph in this lengthy comment in case responses are inserted interstitially. --Moonriddengirl 19:21, 22 August 2007 (UTC))

Comment part 4 Although there are still other factors to consider, with regards to the question of criteria, I don't know that Wikipedia:Lists needs an overhaul. I think Wikipedia:Lists#Criteria for inclusion in lists could stand the addition of maybe a sentence, something like "Review Wikipedia:Lists (stand-alone lists) for further clarification (see also WP:NOT#DIR)." I think Wikipedia:Lists (stand-alone lists) needs to be clarified in terms of why "persons of nationality" counts as "notable for" when, presumably, famous editor John Smith wouldn't be listed on "List of Christians" (even if he does merit a WP page) unless his Christianity has obviously factored into his notability. I think WP:NOT#DIR #1 deserves at least a pointer to Wikipedia:Lists (stand-alone lists) and preferably further definition of "loosely associated," particularly as applies to people. --Moonriddengirl 19:21, 22 August 2007 (UTC)

[edit] Question

So, what's the status of this discussion? Defunct, dead, resolved, gone somewhere else??? FrozenPurpleCube 02:15, 30 August 2007 (UTC)

Good question. :) Right now it's gone quiet. We miss our initiator. I'd love to see it keep going, but have been fairly occupied with other projects. Do you have input? :) --Moonriddengirl 00:48, 30 August 2007 (UTC)

I don't know, I'm pretty iffy on which way to go myself. FrozenPurpleCube 02:15, 30 August 2007 (UTC)

It's a huge issue; the more I try to iron it out in my head, the more it gets lost. :D I'm thinking a simple solution is probably better than a complete overhaul. You've inspired me to run propose my sentence addition to Wikipedia:Lists at that talk page and also at Wikipedia talk:WikiProject Lists. I've once again invited others to join in. Maybe their feedback will spark conversation and forward progress. --Moonriddengirl 12:47, 30 August 2007 (UTC)

Well, I'll check on things there then. FrozenPurpleCube 16:06, 30 August 2007 (UTC)

I hope that they'll come here. I left them a link. :) --Moonriddengirl 16:10, 30 August 2007 (UTC)

• In the meantime, the unwarranted deletion of useful lists continues unabated, pushed by deletionists and approved by hapless closers. See Wikipedia:Articles for deletion/List of Belgian Americans and Wikipedia:Articles for deletion/List of Norwegian Americans for example. Hmains 02:29, 2 September 2007 (UTC)

• Welcome to the conversation. I think many people are genuinely confused about the policies because the policies seem to be conflicted at various locations and have become spread out enough that people may not focus on the big picture. I added a reference at WP:NOT#DIR to refer to Lists (stand alone lists). At Lists I added pointers to both Lists (stand-alone lists) and WP:NOT#DIR. I think somehow a comprehensive understanding of those policies and how they work together must be reached. Right at this moment, one of the lists used as an example in policy is up for deletion. I'm a bit flummoxed. When one policy can be used to condemn an article that another policy implicitly supports, something doesn't seem to be working. --Moonriddengirl 02:49, 2 September 2007 (UTC)

Even worse is that one policy is used by one group of editors to keep a list and the very same policy is used by another group to delete the list. How and why? The policies/guidelines are so loosely written that editors can read them and take them to mean anyting they want Hmains 20:29, 2 September 2007 (UTC)

Is there anyplace where someone is trying to gather a unifying consensus (other than here, of course)? I posted a link to here from an AfD on Wikipedia:Articles_for_deletion/List_of_ice_hockey_players_who_died_young since it seems to have the same problems. Best, --Bfigura ^(talk) 05:35, 2 September 2007 (UTC)

I'm not aware of anywhere else. This page is large enough and the conversation sprawling enough that I begin to think it could be useful to archive it with a summary. Maybe it would help encourage conversation if it weren't so sprawling. :) I may take that on later, unless somebody beats me to it. --Moonriddengirl 11:47, 2 September 2007 (UTC)

[edit] Ah, this is what it was!

So this is the project I bailed on! I almost forgot about it. Wow... this went on for a while, didn't it? Did anything ever come of this?

I'm starting to think it was easier to pass that referendum than it would have been to complete this project... Sidatio (talk) 04:25, 8 January 2008 (UTC)