Wikipedia talk:Good article criteria

From Wikipedia, the free encyclopedia
Jump to navigation Jump to search
ArchiveThis page, a part of the Good article talk page collection, is archived by MiszaBot II. If your discussion was mistakenly archived feel free to go retrieve it.Symbol support vote.svg
Current Archive location: Wikipedia talk:What is a good article?/Archive 3

Template-protected edit request on 10 April 2019[edit]

Please change

:<code>{{Wikipedia:Good article criteria/GAC|1a}}</code> 
:{{Wikipedia:Good article criteria/GAC|1a}}

to {{documentation}}. I have added this to the new documentation page, and this would allow non-template-editors to add categories for the template and otherwise improve its documentation. Thanks, --DannyS712 (talk) 06:08, 10 April 2019 (UTC)

@Bradv: can you take a look? --DannyS712 (talk) 06:08, 10 April 2019 (UTC)
 Donebradv🍁 06:13, 10 April 2019 (UTC)

Potentially reword 0a[edit]

0a, "It is a long way from meeting any one of the six good article criteria" is ambiguous prose IMO. I initially interpreted it as meaning that a quick fail could happen if and only if all six of the criteria were a long way from being satisfied (i.e., "a long way from meeting any of the criteria"), but after looking at a few failed GANs and rereading it I came to interpret it as that if a single criterion was far from being satisfied, which I now think is the correct version.

If the latter interpretation is indeed correct I propose we reword 0a to the effect of "It is a long way from meeting at least one of the six good article criteria", which incidentally also covers articles that are a long way from meeting multiple criteria. John M Wolfson (talk) 03:39, 29 April 2019 (UTC)

Not so sure it is ambiguous, but fine with the change. Not sure it needs the italics though. AIRcorn (talk) 10:36, 29 April 2019 (UTC)
The italics are mine for emphasis (between the changes), my bad. John M Wolfson (talk) 11:17, 29 April 2019 (UTC)
Yeah, just change it from "any" to "at least", and no-one can possibly complain. The Rambling Man (talk) 11:19, 29 April 2019 (UTC)
I agree with AIRcorn and TRM. Best, Barkeep49 (talk) 14:40, 29 April 2019 (UTC)
It doesn't require a change because the original text is not ambiguous. Clearly, "any one of the six" does not mean "any of the six" as you misread it. Chris Troutman (talk) 10:40, 29 April 2019 (UTC)

Template-protected edit request on 28 May 2019[edit]

Template:QF-tags has been moved. Could I please remove it from 0c? 99721829Max (talk) 19:25, 28 May 2019 (UTC)

 DoneJonesey95 (talk) 20:11, 28 May 2019 (UTC)

Understandability criterion[edit]

My proposal to add understandability as a criterion to GA has gained some support (see WP:VPR#Vote). Here are two ways I can think of to implement it:

  1. Add a new section, 1.c, in WP:GACR that reads: "It explains or avoids jargon where possible and is written for as broad an audience as possible". (similar to criterion 6. of WP:BCLASS)
  2. Modify section 1.a in WP:GACR to read: "the prose is clear and concise, the spelling and grammar are correct, and jargon is explained or avoided where possible;"

Which one of these seems preferable? Do you have another idea in mind on how to implement it?

Additionally, Trovatore wanted to make sure that it is specified that the reviewer should either be knowledgeable on the article's topic, or ask for expert help. Mentioning this in the criteria itself seems doesn't seem like a good idea. My idea is to append it at the end of the lead: "The good article criteria are the six standards or tests by which a good article nomination (GAN) may be compared and judged to be a good article (GA). A good article that has met the good article criteria may not have met the criteria for featured articles. Reviewers not familiar with the article's topic are expected to do basic research or ask for expert help."

Thoughts?--Megaman en m (talk) 09:11, 24 July 2019 (UTC)

  • I'd support modifying 1.a., as I think it's best not to create new subcriteria unless absolutely necessary. I'm a little iffy on adding qualifications onto the criteria page itself, perhap suggest appending these to the end of the "Step 3: Reviewing the article" section of the instructions? Iazyges Consermonor Opus meum 20:22, 24 July 2019 (UTC)
    • @Iazyges: I can see technical hurdles to adding new (sub)criteria, namely having to update a bunch of templates like {{GAList}} - is that what you're referring to? Or is it more a WP:CREEP concern? Colin M (talk) 21:56, 25 July 2019 (UTC)

1a proposal[edit]

Participants have come to a consensus that good article criterion 1a should read: The particular style (for example, Czar's proposed wording) and whether to include a footnote can be handled in later discussions. Thank you to everyone who participated; this edit implemented the change. Wug·a·po·des​ 20:19, 1 September 2019 (UTC)
The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Should Good article criterion 1a be changed from the prose is clear and concise, and the spelling and grammar are correct; and to the prose is clear, concise, and understandable to an appropriately broad audience; spelling and grammar are correct? 07:10, 27 July 2019 (UTC)

the prose is clear and concise, technical terms are explained or avoided where possible[1], and the spelling and grammar are correct}}? 21:24, 24 July 2019 (UTC)


  1. ^ While articles on general topics should provide a broadly understandable overview of the topic, articles on specialized topics can assume appropriate background knowledge.


  1. Support -- Iazyges Consermonor Opus meum 17:28, 25 July 2019 (UTC)
  2. Support Mostly for reasons of common sense - being understandable/accessible to people who are likely to read about the subject is absolutely a quality we should require of the articles we label as being among Wikipedia's best. The guidelines WP:MTAU and WP:AUDIENCE show that there's community agreement around the goal of broad accessibility. This is already something I think about when doing GA reviews (and I imagine many others do too), but I think it's important to spell it out in the WP:GACR. Consistency with WP:B? is a happy side effect. I would have a slight preference for adding a new '1c' criterion, but it's looking like that's unlikely to gain consensus support at the moment. Colin M (talk) 21:26, 25 July 2019 (UTC)
  3. Oppose I think that some tweaking of this criteria could be made, but this hasn't had the polishing necessary to make the change quite yet. I think this tweak is too narrow and gets at merely one aspect of what it's trying to accomplish. Given that the GA criterion have been fairly stable for a long time I also worry about upsetting a lot of precedent without firmer understanding of what's replacing it. Best, Barkeep49 (talk) 14:49, 26 July 2019 (UTC)
  4. Support An elegant phrasing that gets the point across.--Megaman en m (talk) 08:34, 27 July 2019 (UTC)
  5. Support. This is a very appropriate standard for Good Articles, and the 27 July phrasing is much better than the phrasing initially suggested on 24 July. P.S. Wugapodes RFCs are often much easier for new arrivals if the RFC statement shows the status-quo version as well as the proposed version. That makes it so much easier to get grounded in what is being discussed, and easier to isolate what would change. I also suggest tossing in a link to WP:Good_article_criteria#Criteria to conveniently view the full criteria. Alsee (talk) 05:22, 28 July 2019 (UTC)
  6. Support as phrased. I consider this part of "clarity" but I consider the distinction warranted, based on the general incomprehensibility of some "walled garden" topic areas on Wikipedia. WP:MTAU is sufficient for the link, if not more specifically WP:GENAUD. czar 16:59, 28 July 2019 (UTC)
  7. Oppose (hopefully only temporarily) unless/until the inappropriate link on "assume appropriate background knowledge" in the footnote is fixed, per my comments below. —David Eppstein (talk) 19:22, 1 August 2019 (UTC)
  8. Oppose in current form; the wikilink on "assume appropriate background knowledge" seems at best irrelevant and redundant with WP:MTAU already being linked. XOR'easter (talk) 23:45, 1 August 2019 (UTC)
  9. Support This covers a problem that is more prevalent that many realize, especially on technical articles. It's not a matter of technical terms......those are covered by the "appropriately" . It's a matter of doing a poor job of presenting/ explaining the topic. North8000 (talk) 11:56, 9 August 2019 (UTC)
  10. Support—pragmatically, this is something I've generally harped on in all my content reviews, and it's pretty important. We should aim for a basic level of comprehension that doesn't require linking away from the article entirely. Der Wohltemperierte Fuchs talk 16:59, 14 August 2019 (UTC)
  11. Support - sometimes it is not obvious what the appropriate level for the article is, but it may well be a good thing to establish. GAN would be a place do do this if it has not been done. · · · Peter Southwood (talk): 09:22, 17 August 2019 (UTC)


  • I like this, but I'm concerned it's a little too specific. Avoiding technical terms is only part of making an article understandable to an appropriately broad audience. For example, see WP:OBVIOUS. The problem in the Thunderbird example there isn't that it uses technical language, but rather that it omits important background information that readers within the article's audience might not be aware of. For this reason, I prefer the more general wording of WP:B?#6, or the proposed 1c above. Colin M (talk) 20:38, 25 July 2019 (UTC)
Also, I disagree that technical language should be omitted entirely, as long as it can be explained in parenthesis. No need to dumb down articles, people are here to learn something new. FunkMonk (talk) 20:53, 25 July 2019 (UTC)
Yeah, I think everyone's on the same page there, it's just hard to convey the nuances and trade-offs with literally just a few words. I agree "avoided where possible" might be a little too strong. Maybe one of the following would be fairer:
Colin M (talk) 21:42, 25 July 2019 (UTC)
I'm in favor of Colin's wording.--Megaman en m (talk) 09:39, 26 July 2019 (UTC)
@Megaman en m: criterion (singular criteria); a criterium is a bike race and hasn't been used in that formulation for...? Good one! ——SerialNumber54129 14:57, 26 July 2019 (UTC)
I never realized that. The singular form of criteria is actually critirium in Dutch, but not English. Changed to criterion.--Megaman en m (talk) 15:36, 26 July 2019 (UTC)
No problem :) I thought you were being ironic about "understandability"! ——SerialNumber54129 16:16, 26 July 2019 (UTC)
  • I've updated the question with Colin M's wording, though I changed the link from WP:AUDIENCE to WP:MTAU which based on the VPProp discussion seems to capture the spirit of "appropriately broad" better than AUDIENCE, though that essay is still linked in the supplemental footnote. If people feel strongly about the link targets that can be discussed as well. Wug·a·po·des​ 07:10, 27 July 2019 (UTC)
    • I'm partial to WP:AUDIENCE for a couple reasons: 1) It's a little more general, in that it's not limited to "technical articles". Again, I like the example in the WP:OBVIOUS section, which uses the first sentence of the article Ford Thunderbird (which most probably wouldn't call a technical article). 2) I personally find it more concise and well-written. Also, I think the "Use of chromatic scales in early Baroque music" example does a good job of capturing the "appropriately broad" spirit.
    On the other hand, WP:MTAU is a guideline, and WP:AUDIENCE is part of an "explanatory supplement" to the MoS, so theoretically the former could be said to have a stronger consensus behind it. Though in practice, I see sections of WP:BETTER (such as WP:AUDIENCE, WP:ASTONISH, WP:TONE) cited a lot more than WP:MTAU. Would be curious to hear what others think. Colin M (talk) 15:53, 27 July 2019 (UTC)
    I have to admit that my usual experience of the {{technical}} tag (which goes along with MTAU) is drive-by tagging of articles on technical topics by editors who do not understand the article and therefore think it is too technical. But because they do not understand the topic of the article, they also do not understand how much background knowledge might be expected of readers of the article and therefore what the appropriate level of writing for the topic should be. More often than not, they end up inappropriately tagging articles that are already written at the appropriate level. I realize that is not the intent with the current proposal, but I think we have to think ahead to how inexperienced GA reviewers might act rather than how we might ideally want them to act. In the case of the "technical" tag, it doesn't help that the tag asks editors to "make it understandable to non-experts" rather than the more careful wording of WP:MTAU to only do this "as far as possible for the widest possible general audience". So we need to be very careful how we write this. In that vein, I don't think the current proposal's choice of Wikipedia:Summary style as the link target for "assume appropriate background knowledge" is a good one. It is likely to lead reviewers to ask for technical articles to be rewritten in summary style, merely because they are technical. Summary style is not about and should not be about making it possible for readers without the appropriate background knowledge to read an article; it solves a different problem (topics that are too big to fit comfortably into a single article so they need to be broken up into smaller pieces). —David Eppstein (talk) 06:26, 28 July 2019 (UTC)
    I've had similar experiences with the {{technical}} tag. Often, the real problem with the article is not that it is full of technical terms, but that it's an infodump by somebody who never learned how to write clearly, worked over by other people who never learned how to write clearly — you know, scientists. An improved version of such an article might be equally "technical", yet actually accessible to a more broad audience. It's like the difference between a theorem expressed in the terse style of a journal article, versus the decompressed version that a mathematician might give in a textbook or a memoir. XOR'easter (talk) 16:45, 28 July 2019 (UTC)
  • An alternative, for parallelism:

    the prose is grammatically correct, clear, concise, and understandable to an appropriately broad audience.[1]

    czar 17:07, 28 July 2019 (UTC)
  • Three days without a comment, do we have a consensus?--Megaman en m (talk) 07:38, 1 August 2019 (UTC)
    Just five days into a thirty-day RfC? I think not. --Redrose64 🌹 (talk) 11:49, 1 August 2019 (UTC)
    Alright, I was not aware that 30 days had to pass.--Megaman en m (talk) 13:25, 1 August 2019 (UTC)
    The only thing significant about 30 days is that it's the period after which a bot will automatically remove the RfC template. See WP:RFCEND: An RfC should last until enough comment has been received that consensus is reached, or until it is apparent it won't be. There is no required minimum or maximum duration. I think rough consensus is looking pretty clear at this point, but since this is a fairly high-impact change, I don't think it would hurt to wait another week. 6 responses is not a great turnout. Colin M (talk) 14:44, 1 August 2019 (UTC)
    I've posted a notice at WT:GA to get wider feedback. Barkeep49 is right that this has some rather wide implications, and I think the more feedback the better so I'd recommend leaving it open for the full 30 days. No deadline and all that. Wug·a·po·des​ 23:23, 3 August 2019 (UTC)
  • If it helps the proposal pass, I'm in favor of dropping the "summary style" link within the footnote. czar 19:48, 3 August 2019 (UTC)
  • Given the feedback from David and XOReaster, I've removed the footnote which is largely redundant with the MTAU link. Hopefully this version covers all the bases, but it's changed a couple times since the RfC started, so I'm pinging all the participants so they can take another look at it and make sure it has consensus. @Iazyges, Colin M, Barkeep49, Megaman en m, Alsee, David Eppstein, and XOR'easter:. Wug·a·po·des​ 23:17, 3 August 2019 (UTC)
    Support current proposal. Iazyges Consermonor Opus meum 23:19, 3 August 2019 (UTC)
    Support current proposal. Alsee (talk) 06:55, 4 August 2019 (UTC)
    Support current proposal. --Megaman en m (talk) 08:33, 4 August 2019 (UTC)
    Support Colin M (talk) 16:31, 4 August 2019 (UTC)
    Support as revised. XOR'easter (talk) 19:13, 6 August 2019 (UTC)
    Support as revised. —David Eppstein (talk) 04:51, 7 August 2019 (UTC)
    It's been 30 days, is this enough for a consensus?--Megaman en m (talk) 08:36, 25 August 2019 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Criterion #2 Doesn't seem to actually require everything to be cited[edit]

The consensus is against the proposed change of #2b from all inline citations are from reliable sources, ... to all prose paragraphs have at least one inline citation, and all inline citations are from reliable sources....

Cunard (talk) 01:19, 22 September 2019 (UTC)

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

I noticed that no-where in the verifiability criterion does it actually state that all the prose within an article must be cited. It only requires a list of citations, that all citations be reliable, that no OR takes place, and no copy-vio or plagiarism. Nowhere does it actually say "all elements of the article must be cited." It's mostly semantic, but I'd suggest updating #2b from all inline citations are from reliable sources, ... to all prose paragraphs have at least one inline citation, and all inline citations are from reliable sources... Iazyges Consermonor Opus meum 17:16, 30 July 2019 (UTC)

  • Support but I don't think that goes far enough. My prior proposal sought to require fact-checking, as the way the criteria reads now one might assume that the presence of citations from reliable sources doesn't require that each citation is checked to ensure all the text is supported. The giant loophole about sources you have access to allows many editors to just affirm that sources have been cited even if the article is a hoax. Chris Troutman (talk) 17:22, 30 July 2019 (UTC)
  • Comment why doesn't WP:V and WP:RS apply to good articles? The Rambling Man (REJOICE!) 17:26, 30 July 2019 (UTC)
  • Per WP:MINREF unless the article contains specific kinds of claims (challenged, quotes, BLP, etc), it is not required to cite any sources. That said, I think it is completely reasonable for GAs to have a higher than minimum standard for sourcing and would be fine clarifying criterion 2. Wug·a·po·des​ 17:32, 30 July 2019 (UTC)
  • Oppose all prose paragraphs have at least one inline citation is completely arbitrary and not based on policy. The current wording of the criterion lists WP:MINREF and quite a few other types of claims for which inline citations are needed. – Finnusertop (talkcontribs) 18:01, 30 July 2019 (UTC)
  • Question is this some hypothetical problem or are there examples of uncited information which couldn't be verified (aka are false) been found in articles? I ask because I spend a considerable amount of my GA review time verifying information and have had the same experience when others have reviewed my nominations. Best, Barkeep49 (talk) 18:22, 30 July 2019 (UTC)
    @Barkeep49: I failed this GAN due to citation problems. Had I not, it would be a GA with lots of factual errors. Chris Troutman (talk) 18:48, 30 July 2019 (UTC)
  • Oppose (in current form) WP:CITEDENSE specifically says there is no "one citation per paragraph" rule. The GA criteria also link to WP:V, which states that "All content must be verifiable." I don't really see the problem here.--Megaman en m (talk) 19:38, 30 July 2019 (UTC)
    @Megaman en m: I have noticed that some editors translate WP:V to mean that a source might exist to verify a claim in an article, not that the source has been cited and that the source does in fact support the content. To them, V only prevents content to which there isn't likely a published source. Couple that with editors that don't check every citation and you have GAs that contain errors. Chris Troutman (talk) 19:47, 30 July 2019 (UTC)
    WP:UNSOURCED still mentions that contentious material needs inline citations. However, I do agree that GA articles should go above bare minimum citation requirements. I just don't think that forcing a one citation per paragraph rule is the right way to go. It might be better to include Wikipedia:When to cite as a wikilink somewhere for example. It states that close paraphrasing should also be cited along with opinions, data and statistics.--Megaman en m (talk) 08:46, 31 July 2019 (UTC)
  • Oppose as written - as others have said, 1 citation per paragraph is arbitrary and has no policy basis. However, I do think the wording could be improved. I've always interpreted 2b as saying I) All inline citations are from reliable sources, and II) Citations are provided for all direct quotations, statistics, published opinion, etc. However, it technically doesn't say II, at least under a literal reading. I think this is just an unfortunate result of trying to keep the wording of each criterion brief and within a single sentence. Colin M (talk) 21:52, 31 July 2019 (UTC)
  • Oppose all prose paragraphs have at least one inline citation is both policy creep and a gamification of the writing process. XOR'easter (talk) 23:47, 1 August 2019 (UTC)
    I think it will generally work out that high-quality articles will have at least one footnote per paragraph, if for no other reason than that's the style of writing that people who hang out here get accustomed to. But setting a numerical citation density as the threshold to pass strikes me as the wrong way to go. It's inventing a number as a substitute for thinking. The "rule of thumb" mentioned below, with its sensible exceptions (e.g., plot summaries that only need to be "sourced" to the thing they are summarizing), seems a better way to approach the question. XOR'easter (talk) 02:35, 2 August 2019 (UTC)
  • Oppose as stated, but I'm open to a less-strict wording. Roughly the same requirement for each paragraph to have at least one citation holds for DYK (supplementary rule D2). It seems strange to me that an article could pass GA and thereby become eligible for DYK but not pass the DYK review because it wasn't well enough cited. On the other hand, the DYK requirement admits exceptions (it states it as a "rule of thumb", and excepts "the lead, plot summaries, and paragraphs which summarize other cited content"). In practice when I review articles for GA I require each claim to be cited (somewhere in the following text), but as for DYK I don't ask for citations at points in the article that summarize later material, and if a single citation suffices for multiple consecutive paragraphs I think it should be ok to have a single footnote at the end of the last paragraph rather than requiring it to be repeated. —David Eppstein (talk) 00:15, 2 August 2019 (UTC)
    David, the problem with that style is that someone only has to add a statement with an inline ref somewhere in one of the middle paragraphs and it becomes completely unclear whether the first paragraph is cited at all, let alone by the ref at the end of the last paragraph. The same applies to some extent to citing a multi-sentence paragraph just once (and yes, I do that like the rest of you); any addition causes problems as people frequently don't think to repeat the existing citation just before the addition. I think the solution is not to do wordsmithing on the criterion but to automate the connection between cited text (of any number of sentences or paragraphs) and the citation, i.e. it's a managed data structure not an unmanaged/unmanageable text-tangle. If only. Till that far-off day, we really ought to cite every sentence, and if that's too hard, then every paragraph. Chiswick Chap (talk) 20:21, 2 August 2019 (UTC)
    Your same objection[1] would apply[1] to any style less strict than a citation[1] for each separate clause[1] of each sentence.[1] We're not going there.[1]David Eppstein (talk) 22:13, 2 August 2019 (UTC)
    Indeed not, but then people are generally happy for sentences supported by multiple refs to have all the refs at the end: less good when there are a dozen refs in a paragraph [all at the end together], so you're overstating your case. Chiswick Chap (talk) 08:44, 3 August 2019 (UTC)
    Do you want to send NBR 224 and 420 Classes to WP:GAR then? --Redrose64 🌹 (talk) 17:08, 3 August 2019 (UTC)
    I don't see any lists of 12 refs all clumped together at the end of any paragraph in that article. I've added a [...] note above in clarification. Chiswick Chap (talk) 17:36, 3 August 2019 (UTC)
  • Strong Oppose The current version is already departs from policy. This would make it even worse. The current problem is that it misstates policy. Policy places a sourcing requirement for the presence of text. It does not place a requirement for the presence of a source. For example, let's say that there is some text which is already sourced to fulfill wp:ver. It may be useful to add a source which has real-world reliability which does not fulfill the trappings of wp:rs.North8000 (talk) 12:04, 9 August 2019 (UTC)
  • Oppose, If the reviewer has any doubts about any statement, they are free to add a {{cn}} template to every statement that they consider needs a reference, and be willing to explain why. Sometimes we need a citation for every statement in a paragraph. Other times it may be overkill, and sometimes it may indicate that the point has not been expressed very clearly. · · · Peter Southwood (talk): 09:36, 17 August 2019 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.


There is today quite a sharp mismatch between the quick-fail criteria on this page and the Template:QF. Presumably the template needs to be updated. Chiswick Chap (talk) 17:10, 31 July 2019 (UTC)