Wikipedia talk:Featured article criteria/Archive 1

Origins of this page?
This page appears to be the only existing policy on featured articles, other than what little is written at the top of the featured article candidates page. Raul654 has cited it a couple times recently to dismiss objection that aren't based on the standards written here. Now that I've taken a good look at it, it seems he's merely citing standards he wrote himself.

I think we could use some actual community discussion of what should constitute a featured article. I personally use some criteria that are not currently represented here, and I don't want them dismissed as illegitimate just because they aren't written here.

I'll start the discussion with the suggestion that we need to separate the questions of "is this an excellent article?" and "should we put this article on the main page?" Various occasions have come up recently where the first answer might be yes while the other is no. Wiki and Wikipedia might be good articles, but we don't need to look like we spend all our time gazing in a mirror. George W. Bush may be good right now, but putting it on the main page during a political campaign would be asking for edit wars. It would be nice to have a way of recognizing such articles. Isomorphic 04:19, 24 Jun 2004 (UTC)


 * I didn't actually write any of this myself (notice the distinct lack of errors in the writing). I copied it from the FAC page when I shortened the introduction a few months back (it was ridiculously long -- 3+ pages). If you check the page history for the FAC, I think it will show that this was written by several people over the course of months. &rarr;Raul654 04:41, Jun 24, 2004 (UTC)
 * My apologies. For some reason that possibility didn't occur to me.  I don't need to look at the history;  I trust you. Isomorphic 00:58, 26 Jun 2004 (UTC)

Is FA status the same as suitability for the main page?
I also think that some topics simply don't work well on the main page. Articles featured on the main page should have some general-interest appeal, because that's what we're presenting to the general public. We want people to read those articles, say "Wow, that was impressive" and start to look for more. Some articles, by the very nature of their topic, won't do that for most people. An article like Gröbner basis, even if it were a brilliant article on the topic, is too technical for a general reader. It would need to wander far off topic in order to provide enough context. I think that that should be a rough guideline - if most readers will need more context than can fit in the article, we probably should reconsider featuring it.

Comments? Isomorphic 04:19, 24 Jun 2004 (UTC)


 * A discussion touching on this very issue is going on right now at Wikipedia talk:Featured article candidates. There seems to be strong consensus that "being featured" (i.e. listed on Feature articles) is a distinct thing from being put on the main page. Thus any topic deemed worthy of an article (not matter how inflammatory, technical, sexually explicit, whatever it is) can be featured. Thus we can reward all writers whatever their field of interest, and show the diversity of WP. Then a separate set of standards (currently applied ad hoc) are used to determine what goes on the main page each day, thus alleviating the concerns you espouse here. Pcb21| Pete 13:08, 24 Jun 2004 (UTC)

I think we should have two categories -- "featured articles" which are about the most comprehensive and accessible about a topic available on the Internet -- and "good articles"; articles that are well written, wikified, formatted and copyedited, are long enough to constitute proper encyclopedia articles (but not necessarily long enough to be featured), do not necessarily have a picture, and are not necessarily accessible to non-experts. Work on such a collection of articles would essentially be the same thing as quality control for an eventual paper/CD/DVD encyclopedia, I guess. Fredrik | talk 09:47, 24 Jun 2004 (UTC)


 * I mentioned this page on Peer review as a good checklist to work from before taking a page to peer review - David Gerard 18:43, 4 Jul 2004 (UTC)

I didn't realize there was a similar discussion going on elsewhere. I will join that. If "featured article" is not, in fact, defined by "article to put on the main page" then most of my concerns go away. Isomorphic 01:03, 26 Jun 2004 (UTC)

Collating material from several writers can show inconsistency in style and content depth. My colleague who is an editor for a leading directory found that it is the hardest task to compile the work of many writers and publish articles with a consistent style. Sometimes it is easy to see that in webpages as here: [url]http://www.targetwoman.com/articles/[/url]

Image status
I've added a word on being sure of the copyright status of images, as this has become an increasingly common objection on WP:FAC - David Gerard 18:43, 4 Jul 2004 (UTC)

Table of contents TOC guidelines
"...have a substantial, but not overwhelming, table of contents (see Section)." I can't find any guidelines on this. The link only tells how a TOC is generated. Please be a little more concrete on what this means as this can be very subjective. Also would like hints on how to create sections and subsections without generating a TOC. I note that some people use a semi-colon (used in definition lists) to create a bold subheading, but it does not always make the subheading bold on some browsers (e.g. Pocket PC IE) Petersam 00:30, 27 Aug 2004 (UTC)
 * Like most things in Wikipedia, it's a judgement call. My rule of thumb is that if it's longer than my 1024x768 screen, it's definitely too big. If you want prior examples -- prior to it's total restructuring, Warsaw uprising had an extremely overwhelming TOC. See this version. Battle of Normandy also had an overwhelming TOC (31 headings/subheadings) -- it has been shortened down now (19 headings/subheadings). &rarr;Raul654 00:39, Aug 27, 2004 (UTC)

What is comprehensiveness?
This article defines "factually accurate" and "well-written" but not "comprehensive." I propose adding something like:


 * Comprehensive: does not omit anything which is currently known and relevant to the subject.

Objections?

Background: a current featured article candidate, Casualties of the conflict in Iraq since 2003, is being challenged on the basis that since the conflict isn't over, or since the article lacks details that aren't currently known (and may well never be knowable), it can't be considered "comprehensive".

It seems silly to expect an article to cover everything that might someday be known. This would prevent featuring a biography of a person who was still alive or even, technically, a history of a country that still exists. Neow 19:45, Oct 14, 2004 (UTC)


 * It would also be silly to expect most articles to cover everything currently known and relevant. That's not "comprehensive", that's "exhaustive". Encyclopedia articles necessarily present a condensed or potted account of most subjects, and therefore they omit things. (NPOV means omitting in a neutral way, it doesn't mean sticking every detail in.) That's why "comprehensive" is so hard to define, I guess.--[[User:Bishonen|Bishonen (talk)]] 21:20, 4 Dec 2004 (UTC)


 * "Comprehensive" can only refer to what we know about something. Since we can not know the future, the fact that an article on a current event lacks that info, means nothing. All that counts is whether or not the article is comprehensive in terms of what is knowable about the subject. It would be silly to de-list an article on a living nation or a person simply due to the fact that we do not know the full details of that subject's entire existence. Only consider blocking a nomination or de-listing a current FA if it lacks important updates. --mav 03:07, 27 Dec 2004 (UTC)

Ineligibility of lists
I proposed List of Ohio county name etymologies as a featured article candidate, but was met with objections as it is a list. That's fine, but should we not have something in the criteria about lists being ineligible so others do not make that mistake? PedanticallySpeaking 16:43, Dec 21, 2004 (UTC)
 * It is already in the criteria. You just have to infer it a bit from them.  The first part is that it needs to be well written prose.  Lists don't meet that requirement, and that is the major thing. The other is that a list is not "Comprehensive: Covers the topic in its entirety; does not omit any major facts or details."  For example the list of name etymologies is not comprehensive without descriptions of them, history, etc.  In summary, lists are metadata, not great writing. I suppose that can be made more explicit, but avoiding instruction creep is good too.  Few people read the criteria as they are. - Taxman 22:42, Dec 21, 2004 (UTC)
 * I agree 100%. &rarr;Raul654 23:00, Dec 21, 2004 (UTC)
 * If it is already in the criteria then it isn't "instruction creep" to make the criteria more clear by saying that explicitly. Let's not make the criteria a puzzle, if we don't want lists as featured articles lets just say so. Paul August &#9742; 03:13, Dec 27, 2004 (UTC)
 * It's still instruction creep if it makes them longer and harder to understand. But if you can think of a way to include that point cleanly, I'd be for it. We can work it out here if you like. - Taxman 03:21, Dec 27, 2004 (UTC)
 * How about in the bulleted list under "A feature article should:" add: "Not be a list." Paul August &#9742; 03:35, Dec 27, 2004 (UTC)

Concern over one of the criteria
"Include references by extensive use of inline references and/or by including a ==References== section (see Cite sources)." I don't think so! Though inline references are good, you still need to add these to the references section. Nothing in Cite sources confirms that this is acceptable, in fact it says the direct opposite (sources must be cited inline and in a references section). For instance (as I pointed out to mav), let's say we have an online article by The Sydney Morning Herald: we add it as a inline reference. Three months later, the story is archived. Say goodbye to the reference!

As a consequence, I'm ammending this. - Ta bu shi da yu 11:05, 6 Feb 2005 (UTC)


 * I've amended your amendment. A stronger statement can be added once MediaWiki gets a feature that will make it easy to add inline references (working something like footnotes in a modern word processor; by adding an inline footnote, its corresponding entry in the ==Notes== section is updated and the inline cite expands from a number when clicked on; ==References== would then be a human-edited list of the major references used to write the article). --mav 04:25, 7 Feb 2005 (UTC)
 * Please revert your change for now until you get consensus for it. Out of courtesy I did not revert myself. I disagree with your changes and feel it should be left closer to the way it was.  Just because there is not an automatic inline citation system yet does not mean they are not important. I'm open to good debate about how it should be written, but I think your changes are too much. - Taxman 13:15, Feb 7, 2005 (UTC)


 * No I will not revert myself. It was Ta bu shi da yu who substantially changed the meaning of this criteria without first seeking consensus on the talk page, not me. My version has the same message but a different emphasis. It says to "Include references. This must be done by using a ==References== section and should be enhanced by the appropriate use of inline citations. If there are few references and the material is uncontroversial, inline citations are often unnecessary." I think my version is an accepetable compromise between what was there before and Ta bu shi da yu's version. --mav 18:36, 7 Feb 2005 (UTC)


 * Source citation is extremely important. It is perhaps the single most important thing which will make WP into a credible reference work. So I believe it is especially important for an FA to be very well referenced. In addition to a list of referenced works, this should include, either extensive inline parenthetical citations, or footnoted citations, which associate particular passages in the text to the references listed in the "References" section. It will be nice when the software makes this easier. But I agree with Taxman,  we shouldn't let software dictate what makes a WP article great. Paul August  &#9742; 16:23, Feb 7, 2005 (UTC)


 * I've been waiting for this feature for a long time. Has any concrete proposal been made, and has a feature request been filed? If not, it should be done as soon as possible. Fredrik | talk 16:58, 7 Feb 2005 (UTC)
 * I filed the request yesterday when this was discussed. bugzilla bug #1484. &rarr;Raul654 18:08, Feb 7, 2005 (UTC)


 * Excellent. I added a comment. Fredrik | talk 18:24, 7 Feb 2005 (UTC)


 * Is this an encyclopaedia or a set of academic papers? If the former, then inline references are inappropriate, if the latter, lets change the name. Filiocht 16:27, Feb 7, 2005 (UTC)
 * Traditional encyclopedias don't have inline references (or many references at all) because it requires too much space. That is not a concern for Wikipedia. There's no reason not to aim for the highest possible standard. Fredrik | talk 16:58, 7 Feb 2005 (UTC)
 * Traditional encyclopedias also don't have inline references because the vast majority of readers of a print encyclopedia will not use them. They do not adhere to a "lower standard" than academic papers, they have a completely different audience and purpose.  Like print encyclopedias, Wikipedia is aimed at a general audience.  We should use inline citations only so far as they don't detract from readability and ease of use, which are more important than making academics feel at home. Isomorphic 17:23, 7 Feb 2005 (UTC)
 * By using notes, which add minimal clutter, we can aim Wikipedia at both academics and non-academics. Fredrik | talk 17:32, 7 Feb 2005 (UTC)
 * It is not a question of making "academics feel at home". It is a question, among other things, of credibility. Unlike Britannica say, Wikipedia can't rely for its credibility, on the (external) authority and reputation of its editors, nor has it stood the "test of time". As currently conceived, any credibility Wikipedia is to have, will have to come from its (internal) content. To be credible, that content should be easily verifiable. To be verifiable, that content should include extensive references. To be easily verifiable, that content should also include extensive citations of those references, which associate specific parts of the content with specific parts of the references cited. As for readability, I think there is a strong analogy here with regard to links. I also find links visually distracting, but this is more than made up for by the information they provide. Paul August &#9742; 21:47, Feb 7, 2005 (UTC)
 * In short, the demand for inline references stems from a failure of trust in the Wikipedia model. Other encyclopaedias do not use inline references because encyclopaedias are reference works. Filiocht 08:27, Feb 8, 2005 (UTC)
 * What are you arguing for? You keep going around disagreeing with nearly every attempt I make to try to increase the referencing and reliability of Wikipedia. What is anyone that is arguing against the need for inline citations arguing for?  What benefit to the project or to mankind is there to not promoting the need for better referencing?  I can't think of any, but I can think of huge benefits to have a well researched and cited information source that is available for free to everyone. If current trends continue, Wikipedia will soon be the premier source of reference information. It may already be the premier source of free reference information.  Because of that, we have a duty to promote systems and policies that make that information as valuable and reliable as possible.  It is always easier to tear something down (which is what I can't help but feel evryone arguing against references is doing) than it is to build something great. To build something great and remarkable, you have to do something different from what others do, not be held back and limited by what they don't do. Wikipedia is already doing that by being free and free to edit, but to really make an impact it needs to be demonstrably reliable. Only better referencing, including inline citations to reliable sources, can do that. - Taxman 14:14, Feb 8, 2005 (UTC)

Please take discussion and criticism of Cite sources to Wikipedia talk:Cite sources. Thanks. Hyacinth 00:29, 8 Feb 2005 (UTC)
 * What I'm arguing for is a level of referencing that will meet the needs of the average reader. This person is the one who is most often overlooked by Wikipedians, especially Wikipedians on a mission. Over-referencing gets in the way of most people who use an encyclopaedia (experts, by definition, will not be general readers) and it is a waste of resources, as well as a futile pipedream, to insist that all articles are referenced to a level above user needs. I work every day in a world where exact fact-by-fact referencing system is vital and that's fine in that environment, but it just is not appropriate to a general reference work. By the way, I would say that the onus is not for me to say what I am for but for you to justify overriding the principle of parsimony. KISS is a pretty good basis to go by for a project aimed at the widest possible readership, insisting on a level of complexity that will put genuine contributors off is not. Filiocht 14:24, Feb 8, 2005 (UTC)


 * Mate, we aren't talking about the normal articles. We are talking about featured articles! IMO, if you want your article featured, you'll spend a (very short) amount of time referencing the material. Referencing material actually is beneficial, because then we can stop subtle vandalism. For instance, someone changing the birth date or adding birth dates that are plainly wrong. Take, for instance, the GNAA, who have a script that adds subtle changes that are wrong. A while ago a newspaper did the same thing, to (of all things) Ecco the dolphin article. That's not a controversial article. Yet by asking for references (and again I'm diverting from the fact that this is a "What is a featured article") it would have stopped them in their tracks. Now if we say that it's not really necessary, then they'll just say "this isn't a controversial article!" So there's another issue we need to address. What is "controversial"? For instance, is bash.org a controversial article? To the average user, no it's not. However, the fact remains that it had a totallydisputed tag on it at one point! Do you see my point? How do we define what "controversial" means? Who determines this? Does someone have to kick up a big stink about the fact that they think that normal material is dubious to do that? - Ta bu shi da yu 21:46, 8 Feb 2005 (UTC)


 * If someone I don't recognize changes dates on an article on my Watchlist and the edit seemed fishy (i.e., that was the only edit or there is a series of similar edits), I would either revert immediately if it was an anon or request verification if it was a logged in user (or maybe revert first and then also ask). I'd hope most editors who've been around awhile would do the same. That has little if anything to do with requiring citations up front for every mundane detail in anticipation of the possibility that it might someday be vandalized in a subtle way. As for what is controversial, no, I don't see your point. An article or a detail within an article can be the subject of dispute without being a controversial topic. There is no necessary correlation between a controversial topic and disputing the factuality of an article. What makes something disputed is that someone challenges something. If someone is simply being irrational and persist without articulating why or what they are disputing, then the situation probably needs to be escalated in the dispute resolution process. older &ne; wiser 00:35, Feb 9, 2005 (UTC)


 * Filiocht has expressed many of my concerns regarding (over)-referencing very well. IMO, much of this hand-wringing over referencing and fact-checking is overkill. Editors should be encouraged to provide sources for material that they add when it is derived from those sources. However, specific citations should really only be needed for controversial or disputed content. For the purposes of a general encyclopedia, a Bibliography or Further Reading section is more helpful than a long listing of specific sources for mundane facts. older &ne; wiser 15:08, Feb 8, 2005 (UTC)


 * I think I have laid out the value of doing it very clearly above. You have not mentioned any important costs to better referencing, and even if there were, they would be well outweighed by the massive benefits. That very simply justifies setting aside your beloved Occam's razor. Justifying that better citations are not needed only by Occam's razor is aiming toward mediocrity. If the system is put in place, with the proper motivations, Wikipedia could eventually be the most reliable resource available, and be free. Again, what benefit is there to arguing against that? All you have said is that it is not needed, but I don't think you can come up with any actual benefits, and again, certainly not any that would outweigh the value. You say it gets in the way. Mediawiki markup could in a very short relative amount of time come up with an inline referencing system that could even be turned off by default, so those that don't care are not even burdened a little by it. Cost eliminated. I think I've shown by the above that the value alone makes it not a "waste of resources", but don't forget that Wikipedia editors are volunteers, making their cost zero beyond server and bandwidth. So we are left with you arguing against something that has great value, just because you feel it is not needed. Which, as already mentioned, amounts to tearing something down instead of contributing to making it remarkable. There is no justifiable reason to limit the quality of the project to no better than what paper encyclopedias do, when it has the capability to do much better, and the value to doing much better is so great. Lots of things were originally pipe dreams that make your life better every day. Bkonrad, there is no hand-wringing or worrying, simply working toward a system that makes Wikipedia more valuable. I agree Wikipedia doesn't "need" citation for mundane facts (In fact Wikipedia doesn't need anything). But I do not believe it is bad if they are. Part of that is because what is mundane to you is dubious to someone else. - Taxman 17:16, Feb 8, 2005 (UTC)


 * Justifying only by Occam's razor is aiming toward mediocrity. I think you are misunderstanding Ockham's razor then. Although you presume that you have already proven your points, I beg to differ. The specifics have not been thoroughly discussed outside of a few corners of Wikipedia, and even in those places I have not seen complete agreement about what is the best solution. It's one thing to talk about vaporware or dreamware and another thing to specify exactly how it would actually work and interact with everything else. Further, while you presume that such a citation system would solve all problems of reliability--in fact it only adds another layer of complexity. Do you really think that POV trolls and vandals would not be able to figure out how to game such a system? I am not opposed to implementing some way to make citing sources easier. But I very much disagree with some of the specific approaches being advocated. IMO, the best approach is to use the simplest solution and not go overboard with adding unecessary complexity into an already complex and highly dynamic system. If additional complexity is truly needed, then we need to proceed with great care to ensure that we do not fall afoul of unintended consequences. older &ne; wiser 00:35, Feb 9, 2005 (UTC)
 * Ahh, well that was not what I meant. I was unclear there, so I adjusted my comment a bit. Filiocht, if I get him right was using Occam's razor as saying that did not justify the extra complication of having more/requiring inline citations at all. I certainly agree the simplest inline citation system that meets the most people's needs is the one that should be used. But to say that inline citations should not be done because they are not needed is flawed. - Taxman 00:47, Feb 9, 2005 (UTC)
 * I have no problem with providing references; I do it all the time. I do have a problem with the clamour for inline references. The only benefit is that rather than having to buy or borrow the book in question and look up the index, our putative fact-checker will only have to go direct to the cited page after acquiring the book. And it is perfectly possible to cheat this system on the plausible assumption that adding references of this sort make fact-checking less likely because of the air of accuracy they add. And they are in the way of the general reader. And if the original author fails to add them the article will never have them and therefore can never be good enough for some measures of what is acceptable quality. Should such articles be deleted? (I am now taking this to its logical conclusion beyond the FA debate).
 * By the way, print encyclopaedias do not provide inline refs, not because of issues of space but rather because they are considered to be works of reference in themselves. I still feel that Occam's Razor is a good criterion here; does the benefit of asking authors to provide inline refs rather than a comprehensive list of references at the foot of the article justify the extra layer of complexity added. My own view on this is, I hope, clear. Filiocht 08:55, Feb 9, 2005 (UTC)


 * And if the original author fails to add them the article will never have them and therefore can never be good enough for some measures of what is acceptable quality.


 * Anyone can add sources later on. It's just a lot more work. This is not significantly different from text wikification.


 * Finding out which books were used is a lot more work. Adding inline refs to specific pages (the call that I am opposing) is hugely more work. If you do not believe me, pick an article and try it. Adding references does not mean adding any old source on the topic, it means adding the exact references used by the original author. So this problem does not just go away, you know. Filiocht 09:58, Feb 9, 2005 (UTC)


 * Why would you have to add the exact source used by the original author? If you really do, just remove the fact and then reinsert it with your own reference. Fredrik | talk 10:37, 9 Feb 2005 (UTC)


 * By the way, print encyclopaedias do not provide inline refs, not because of issues of space but rather because they are considered to be works of reference in themselves.


 * I don't think lack of sources is an inherent property of "reference" works at all, just a misconception you've been comforted into. And if it really is the correct definition, I see no problem with setting the goal for Wikipedia higher than being just a "reference" work.


 * Also, Wikipedia is not a finished encyclopedia sitting on a bookshelf somewhere; it is a work in progress and subject to an editing process that would benefit tremendously from inline references. If you'd like to make a "stable" release or print, removing existing references would be the easiest thing in the world. Fredrik | talk 09:39, 9 Feb 2005 (UTC)

Citation required?
Here is the original text:
 * "Include references. This must be done by using a ==References== section and may be enhanced by the appropriate use of inline citations. If there are few references and the material is uncontroversial, inline citations are often unnecessary (section then listed as ==Further reading==), though strongly advised (see Wikipedia:Cite sources)."

Here is my clarification:
 * "Include references or further reading. This must be done by using a ==References== section enhanced by the appropriate use of inline citations. If there are few references and the material is uncontroversial, inline citations are often unnecessary (though strongly advised, see Wikipedia:Cite sources) the section may then be listed as ==Further reading==."

Here is Maveric149's "wtf" version:
 * "Include references. This must be done by using a ==References== section and may be enhanced by the appropriate use of inline citations. If there are few references and the material is uncontroversial, inline citations are often unnecessary, though strongly advised (see Wikipedia:Cite sources)."

Hyacinth 01:10, 16 Mar 2005 (UTC)
 * Sorry, when I saw the diffs I thought you had added the part about further reading. That part is completely innapropriate, making either of the first two versions above unnaceptable. Further reading is an ambiguous term that could be used to list works not used at all as a reference, but just there for more material for an interested reader. As discussed ad nauseum in earlier sections on this page, that is horribly unhelpful if references are what is needed. - Taxman 04:06, Mar 16, 2005 (UTC)