Wikipedia talk:What the Good article criteria are not

src: www.slate.com

Video Wikipedia talk:What the Good article criteria are not

Suggestion

All too often, essays don't catch up with the changes in the relevant policies. Many are simply abandoned and left to rust while the rest of wikipedia goes their own way... Every time the reader needs to check and compare the edit histories (and wow to those who don't). Right now this is not the case (the essay was written a few days ago and there were no major changes in WP:GAC afterwards) - but who knows, right? Remember the MOS wars.

I'd suggest that this and similar essays always carry a disclaimer saying that "This revision has been checked for consistency with governing policies and guidelines as of (date)." East of Borschov 19:07, 8 August 2010 (UTC)

This is a good idea, but note also that the essay actually transcludes rather than quotes the GA criteria. Geometry guy 21:02, 8 August 2010 (UTC)

Thanks for your comments.

That concern is exactly why I transcluded the GA criteria. I didn't want the maintenance hassle involved in minor changes, and if there were major changes, I wanted the mismatch to be right in front of any editor who had his/her eyes open.

I think that the only significant unaddressed risk is if we added a new criterion. WhatamIdoing (talk) 07:06, 9 August 2010 (UTC)

Maps Wikipedia talk:What the Good article criteria are not

Stability

I think the essay understates stability:

• Merge or splitting articles between nomination and start of review would be confusing, and merge or splitting during review would be worse.
• As either as nomination or review, I'd avoid major changes of the article once the review starts, and instead summarise them as comments for discussion during the review. --Philcha (talk) 21:50, 8 August 2010 (UTC)

Thanks for your comments.

For merges, it really depends on the nature of the merge: Many 'merges' look like redirecting the other page, and doing nothing (or very little) to the target. That sort of merge doesn't harm the page stability and shouldn't interfere with the GA review.

For splits (and more substantial merges), I agree that an agreement to actually split (or merge) the page is usually a serious blow to stability, but merely proposing it is not (especially if the editor who proposed it is being told that it's a lousy idea). WhatamIdoing (talk) 07:01, 9 August 2010 (UTC)

src: upload.wikimedia.org

Inline citations

This is a great essay by the way, I've already found it very helpful. I have a slight quibble, but maybe this isn't the right place. I'm not sure if I'm at odds with the essay or the actual criteria, so feel free to pint me elsewhere.

I feel that for "factually accurate and verifiable", the lack of a requirement for inline citations is overstated. In my understanding of WP:V, all info must have come from somewhere, unless it is blindingly obvious or undeniably easy for the reader to verify themselves. Saying that we don't need inline citations makes the assumption that the info used in the article is sourced from one of the sources listed at the end. This is very often not the case, particularly with articles that rely heavily or solely on online sources and where no general sources are used. Very often in articles I've come across at GAN, if it's not cited inline, it's not sourced at all.

"Counter-intuitive or controversial statements that are challenged or likely to be challenged" require inline cites, but what about statements that aren't really controversial, but that I as a reader have no idea whether or not they're true, and no idea where to find the source? Ok, maybe this is more directly related to the criteria themselves, I just feel that the essay is overstating it a bit, and discouraging reviewers from asking for citations/sources (sometimes the same thing) that should actually be there.--BelovedFreak 10:20, 9 August 2010 (UTC)

Yes, I think your concern is with the actual criteria, which are fairly limited.

Do you generally read all (or nearly all) of the sources named in an article (and perhaps a few more, so you can make sure that there isn't some unfortunate cherry-picking going on)? If so, that step should let you know whether a non-controversial fact is verifiable -- even if there are zero inline citations in the article. You can reasonably treat all inline citations as also being general citations.

The question I think you need to be asking yourself is this: When you're certain the non-controversial material is not required to have an inline citation (by GACR or any content policy), then how do you conclude that it "should" have one? WhatamIdoing (talk) 16:29, 9 August 2010 (UTC)

In answer to your question, yes I do read the sources. Most of the articles I have reviewed have mostly relied on online sources, so that is easy enough. I am talking about information not found in the sources cited. I suppose I have been labouring under the delusion that we were supposed to point the reader to where they could verify the information. Anyway, I don't think I'm expressing what I mean very well, and it's clearly irrelevant to this essay, so I'll leave it there. Thanks for your reply.--BelovedFreak 20:07, 9 August 2010 (UTC)

Wikipedia should not contain material which cannot be reliably sourced. The GA criteria demand references (i.e., a bibliography) to all sources of information used in the article. But the GA criteria only ask for inline citations for material which the reader is most likely to want to check. Beyond that, many articles have only a few key secondary sources, which generally support uncontroversial claims made in the article, and these can be listed as sources without citing them regularly for every detail.

This essay is primarily written to counter reviewers asking for more than is required by the GA criteria (and requiring a certain amount of inline citation density has been a problem in the past). Consequently, there is less emphasis in the essay on issues where reviewers may ask too little. If that is a problem, it could be made the topic of another essay! Geometry guy 20:22, 9 August 2010 (UTC)

Ok, you're right. You know, I'm not actually disagreeing with this essay, or the criteria, or what WhatamIdoing said above, but I think I'm just looking at this essay in the wrong way.--BelovedFreak 21:24, 9 August 2010 (UTC)

Belovedfreak, I want to thank you for looking at this essay in the "wrong" way, and being willing to share your reaction. It's actually very helpful.

Geometry guy, I've been mulling over this question for a while, and staring at the criteria, and I think the answer is "no". That is, the GACR as currently written do not seem to require any sources in the article for any fact except for the five types of statements that require inline sources. They must be verifiable (e.g., at your favorite web search engine), but not verified. So, e.g., if an article says that Bald Mountain is south of Hell (pretend my example is True(TM) ;-), and zero of the sources cited in the article say anything about the location of the mountain -- but you can find a RS at your FWSE that supports this, or you happen to know it's factually accurate -- then this uncited-but-verifiable statement complies with the actual sourcing criteria.

However, even if an article does not contain any of the five kinds of material that require an inline source, then a completely uncited article couldn't be declared a Good article (even if all facts are accurate and verifiable), because 3a requires at least one level 2 section that names at least one reliable source, and thus implicitly bans unref'd articles.

Is that how you (and anyone else who's watching) read it, as well? If so, then perhaps we should add a sentence about the difference between "verifiable" and "cited". WhatamIdoing (talk) 22:16, 9 August 2010 (UTC)

2a: "it provides references to all sources of information in the section(s) dedicated to the attribution of these sources according to the guide to layout" (my emphasis). Geometry guy 22:34, 9 August 2010 (UTC)

I hadn't thought of looking at it that way. (This is why we have these discussions.  ;-)

But was any reliable source actually used? What if I have added the information about the location of Bald Mountain from my own personal experience/expert knowledge/recollection of what someone once told me? (It's not a WP:NOR violation because I have specified that the statement is factually accurate and verifiable.) If I used no reliable source in writing that sentence, then should a reliable source be provided for it? WhatamIdoing (talk) 23:13, 9 August 2010 (UTC)

(<-) A big problem with WP:V is its ill-defined notion of "challenged" and "likely to be challenged". In the context of a review it could be interpreted as meaning that any request for inline citation by the reviewer is a challenge (making the issue 100% likely to be challenged, hence requiring an inline citation). Criterion 2 discourages such an interpretation, because it is essentially vacuous and thus completely unhelpful. Instead it asks for more sourcing than WP:V requires and sets a minimum benchmark for inline citation. If no sources were used for a particular fact, as in your scenario, then there are two possibilities:

• The fact is common knowledge in the context.
• It isn't.

In the latter case, your recollection of what someone told you once is not a reliable source (sorry about that :) ), and the GA criteria require that the article makes reference to better one(s), though not necessarily through inline citation. Geometry guy 20:46, 10 August 2010 (UTC)

So are you exempt from the need to WP:SAYWHEREYOUGOTIT if you got the verifiable information from an unreliable source? WhatamIdoing (talk) 23:26, 10 August 2010 (UTC)

No, according to that guideline you have to refer to both the original source and your source. So the references and/or the inline citation (if there is one) would say "Scholar (2006), as cited by Friend In Pub (2010)". At which point, the GAN reviewer might reasonably ask that you check the original source. Geometry guy 21:54, 23 August 2010 (UTC)

Imagine that there's no good source underlying it, e.g., I learned from a random blog that Bald Mountain is south of Hell, and the blog doesn't name any other sources. It happens that it's True(TM), but I haven't used a reliable source to write this.

What would you do? Leave it uncited (unless/until a WP:BURDEN challenge appears), since I actually used no reliable source? Say that I got it from an unreliable source? Cite something that you didn't actually use? Deliberately 'use' a reliable source just so that you can cite something that isn't embarassing? WhatamIdoing (talk) 21:55, 24 August 2010 (UTC)

GG, you can't have recommendations here that violate the content policies. V requires inline citations. We have to make that clear. [1] SlimVirgin ^{talk|contribs} 22:22, 29 September 2010 (UTC)

You're reverted my compromise too. [2] If you want to argue against the policy, you'll have to do it there, but that part of it (in fact all of it) has strong support. You can't create an essay that makes GA nominees think it's okay to ignore the sourcing policy. SlimVirgin ^{talk|contribs} 23:04, 29 September 2010 (UTC)

I'm just chipping in here on the offchance you are referring to me. I am not the creator of this essay, nor do I respond to the handle "GG". Thanks, Geometry guy 23:09, 29 September 2010 (UTC)

I think WhatamIdoing has created a wonderful and much needed guide, and any flaws in this essay are due to the GA criteria rather than the writing of the essay. In the Factually accurate and verifiable section the criteria itself is "A good article is-- ... Factually accurate and verifiable", with a wikilink to Wikipedia:Verifiability, one of our core policies. That link is important. ALL articles are subject to that policy, Good or otherwise. The sub-sections aim to point out the key elements of that policy to assist writers and reviewers of Good Articles. If the sub-sections are at variance with the policy, then the policy takes precedence. The policy states: "anything challenged or likely to be challenged, including all quotations, be attributed to a reliable source in the form of an inline citation, and that the source directly supports the material in question." That is rather easier to follow and understand than the current GA criteria explanation of "counter-intuitive or controversial statements that are challenged or likely to be challenged". We need to change the wording of the criteria from "counter-intuitive or controversial statements" to "any statements".

Looking closely at the history, the change of the wording from policy was made Sept 2007 by User:Deckiller after discussions here, here, here, here, here, here, and here. The move to reduce the wording from following policy occurs in A proposal, and is generated by personal feelings against the implications of the policy wording: "I long for the day when "likely to be challenged" is removed from our collective vocabulary. It's simply an excuse for Geogre-ish non-referencing" / "I find that spotty as well. Though getting rid of that might be overreaching for now." / "Why not? We may as well clear up any ambiguity right here and now. I could think of better ways to word it, perhaps "extraordinary claims" (Bob is an alien needs citation) or "hard statistics and facts," or "information culled from experts" (which is highly unlikely to be common knowledge)." / etc.

The general improvements made as a result of those long discussions was very good, but the final wording got side-tracked by the personal opinions right at the end, and people lost perspective that what they didn't like was actually policy wording, and changes to policy wording should take place on the policy page. The wording before the change was "material that is challenged or likely to be challenged".

While understanding and respecting people's aversion to the wider implications of the policy wording, the change to "counter-intuitive or controversial statements" was inappropriate and should now be changed back to "any statements" or "any material". SilkTork *^YES! 11:01, 30 September 2010 (UTC)

I tried to be bold and amend the criteria, but my edit hasn't worked. Probably a good thing, as it should be discussed first. SilkTork *^YES! 11:16, 30 September 2010 (UTC)

Hi SilkTork and thanks for commenting here. I agree with you that Whatamidoing has done an excellent job with this essay, which responds to a tendency for reviewers to impose their own criteria on an article ("my way or the highway" is one description I have seen of this).

Your edit to the GA criteria worked, but won't feed through until the GA criteria page is purged. However, I agree with you that changes to the criteria should be discussed first at WT:WIAGA, so I have reverted your edit. While I sympathise with your motivation to make this change, I disagree with it, so let me explain.

The GA criteria are not policy like WP:V. The do not and cannot overrule policy. Indeed as far as I am aware, they do not even have official status as a guideline. They are not part of the encyclopedia, but are instead part of a (very well established and successful) behind the scenes process to improve the encyclopedia, known as GA. Consequently they have been adapted over many years to optimize that process.

I have no problem with WP:V stating as policy that any material that is "challenged or likely to be challenged" should have an inline citation. It is a grand statement that reflects our vision for an encyclopedia that readers can trust. Unfortunately it is also a rather ambiguous phrase (challenged by whom? how likely?). Consequently, even though the GA criteria link to WP:V, it is appropriate for them to give a more precise "GA meaning" to the phrase rather than quoting it directly. Otherwise, reviewers could simply say "I challenge this statement" and hence impose their own preference for inline citations (and inline citation density) on any article they review, even one with just one substantial secondary source.

The history of GA is very interesting, as initially it was rather lax about citations, but then there was a huge crackdown which was very unpopular. The discussions you refer to from 2007 were a response to that, and a step back from an extreme position. I think the current criteria provide a good balance of specific advice to reviewers about inline citations. I have been encouraging stable criteria and moderate changes since that time. If you believe that the balance needs to be shifted, please open up a thread at WT:WIAGA. Geometry guy 22:33, 30 September 2010 (UTC)

I think that one of the mental shifts necessary to understand GA is that it's not a "mini FA". We're not looking for near-perfection. If you don't provide an inline citation for non-controversial material (but the sources do exist, and thus the material is verifiable), then that's okay for GA. People shouldn't be failing GAs because there's no citation after "The human hand normally has four fingers and one thumb," or "Arizona is in the southwest United States," or "Elvis Presley was a famous musician." WhatamIdoing (talk) 04:41, 1 October 2010 (UTC)

You might be amused by this diff. However, it's a good illustration of the problem with WP:V as it's currently written. As Gguy implies, not every challenge is predictable or valid. For me, interpretation of "challenged or likely to be challenged" generally boils down to I know it when I see it... which is usually fine for experienced editors that are well-versed in article writing norms but not so useful for new or inexperienced editors. Personally I'd like to see a clarification along the lines of "challenged or likely to be challenged by a reasonable, educated layperson", but that discussion's for another place and time :) EyeSerene^talk 08:52, 1 October 2010 (UTC)

Personally, I'd address the WP:V issue by changing it to "challenged". Good editors will normally provide citations when they write material, just so they don't have to mess with it later, but "required when challenged, and not technically required otherwise" eliminates all the guesswork. WhatamIdoing (talk) 02:10, 2 October 2010 (UTC)

I love the 5 fingers example!

Changing WP:V simply to "challenged" would be a significant shift which many editors would oppose. We have to accept that there are differing opinions, editors with different backgrounds, and different interpretations within different fields as to what "verifiability" means, and different problems faced by different articles: "likely to be challenged" covers these gray areas. But at the level of a GA review, even "challenged" is problematic, as the reviewer (or any other editor) can challenge any statement they like by adding a "citation needed" tag. Indeed, we want reviewers to make legitimate challenges. But what is a legitimate challenge?

The present criterion is a modus vivendi and operandi that has worked well at GA for many years. It is a delicate balance between reviewer independence and review consistency, but has worked well in terms of low levels of disputes and reassessments since 2007. Geometry guy 21:09, 2 October 2010 (UTC)

It made me chuckle when I saw it :)

I agree, GA has been largely successful in furnishing its reviewers with a pragmatic rule of thumb that can be applied to produce a decent standard of verification while avoiding plumbing the depths of the wikilawyer's charter that WP:V has the potential to become. Although I understand where the objections are coming from, I don't accept that GA sourcing guidance is at odds with WP:V. It simply attempts to define a realistically achievable subset of the wider sourcing criteria. EyeSerene^talk 09:32, 4 October 2010 (UTC)

Obviously the issue here is with the criteria and not reviewing. And when reviewing one should not set one's own policies. That said, I think the project should "bite the bullet" and require inline citing. And set a citation style. Saying "well you can do it other ways" is theoretically OK, but in practice leads to very poor quality control (I can prove it by random article visits). It's sort of a historic remnant of the looser wiki. But we should just get over it and upgrade. TCO (talk) 22:20, 24 January 2011 (UTC)

src: upload.wikimedia.org

Likely

Under "Factually accurate and verifiable", what do you think about linking to Wikipedia:Likely to be challenged?

We occasionally seem to have editors interpret "likely to be challenged" as meaning something closer to "conceivably possible that someone might eventually challenge this". WhatamIdoing (talk) 21:06, 25 November 2010 (UTC)

src: upload.wikimedia.org

Great essay

I always push people when thinking about corporate strategy to better define what they won't do, what they won't emphasize and to clearly articulate areas of trade-off (i.e. we will work on this and reduce effort on that). REally like the "what it is not" thought experiment here.TCO (talk) 22:21, 24 January 2011 (UTC)

src: upload.wikimedia.org

Accepting sources in good faith

About the recent addition "Sources should not be accepted in good faith":

I don't know if this quite hits the right note.

It's absolutely true that a reviewer should not completely skip the citation review stage with lazy handwaving about "good faith". Reviewers should be aware that, with the best of intentions, the editors might have made a mistake, and it's the reviewer's duty to find any such mistakes now, to save the authors any embarrassment if they decide to pursue FA.

But I don't think we are requiring scrutiny of 100% of sources. I think it acceptable to pass an article on the assumption that a few sources were probably used correctly, at least if they don't cover contentious BLP matter and your review of the majority of sources has demonstrated that the editors have generally been careful. WhatamIdoing (talk) 17:16, 1 March 2011 (UTC)

src: upload.wikimedia.org

Elaboration of a mistake to avoid

I'm looking for some elaboration of this mistake to avoid: "Requiring the inclusion of information that is not known or addressed by reliable sources."

How specific is this intended to be? Does "information" mean "a specific fact", or does it go all the way up to "all the major aspects"? The reason I ask is that I initiated a good article reassessment of Wait Your Turn based on its dearth of information about the song's writing, recording and composition, which I believe are the major aspects of a song article. The participants arguing against delisting have stated that the lack of coverage in the article does not preclude its GA status because such information is not available, which I believe they got from this essay. But it doesn't seem like this one mistake to avoid was intended to trump the actual criterion (#3a). Two Hearted River (paddle / fish) 03:29, 25 August 2011 (UTC)

src: upload.wikimedia.org

Reviewers' hobby-horses

It seems to me that a long, detailed article may not be suitable for GA review, because any reviewer is likely to disagree with some aspects of such an article. I am having a drawn out discussion with the GA reviewer of David Hume and I'm thinking of withdrawing my nomination of what I think is a GA. For instance, this reviewer objects strongly to the number of quotation marks in the article. S/he also thinks that some material, although present, should be put in another part of the article, although previous editors have left it where it is. Furthermore, s/he thinks that there is not enough emphasis on the financial situation of Hume, even though there is a line that points this out, and previous editors have not thought it very important. This seems to me to go way beyond the GA review guidelines. I even had to argue against the idea that the names of writers of RS references should not be mentioned in the text unless they were famous and had their own WP article - this objection was withdrawn. As WhatamIdoing points out above, "GA is ... not a "mini FA". We're not looking for near-perfection". Perhaps this should be emphasized more in the review guidelines. Myrvin (talk) 13:11, 1 December 2014 (UTC)

I'm sorry you're having such a frustrating time. Some reviewers are better than others, just like some noms are better than others. You have the option of withdrawing it and re-nominating it later, if you want.

In general, my suggestion is for you to pick one issue at a time, and ask the reviewer to specifically identify, by number, the exact criteria that the objection is based on. Sometimes this helps the reviewer get away from "not how I would have written it" to "what the criteria say". If a plausible criteria is identified, then you can ask for a second opinion on the specific question of whether (for example) due weight has been given to Hume's financial situation. Good luck, WhatamIdoing (talk) 08:58, 3 December 2014 (UTC)

Thanks very much for the advice. What has happened is that another reviewer has volunteered to take over. The original (actually the second) reviewer has failed the article, and I've put it into the system again. So, third time lucky perhaps. I think the "too many quotations" was the main problem. I can't find any rules about this. The reviewer said they made the article's language less than clear and concise. Myrvin (talk) 10:40, 3 December 2014 (UTC)

WhatamIdoing: I think there is a problem with the "clear and concise" requirement. Any reviewer who doesn't like the way an article is written - say with "too many quotation marks" - can simply say that, because of this, the article is not clear and concise. This could happen with any MOS entry that the reviewer likes, or just because of a hobby-horse, with no MOS justification. I wonder if this catch-all GAC could be made clearer.

Also, I have fallen foul of a reviewer's liking for a particular aspect of an article that other editors seem to think of little importance, so have ignored. In the case of David Hume, the reviewer thought that his financial situation was very important. I pointed out that this is not a good enough reason to fail the article, since GAs do not need to include every single aspect of what an article could cover. This cut no ice, even after I included a line to say he needed to work for a living. It now turns out that what s/he really wanted was a thorough re-write. Very frustrating for a nominator. Myrvin (talk) 09:42, 5 December 2014 (UTC)

"Clear and concise" can be a problem, although I think that "well-written" is probably more open to abuse. With "clear", you can ask whether the reviewer understood the sentence? If so, then it's clear (Right? Because if it was unclear, then the reviewer would not understand it). With "concise", you can ask whether the reviewer believes that all this information could be conveyed in a smaller number of words. But with "well-written", it's pretty subjective.

I hear what you say about hobby horses. A few years ago, I saw one review of some award in which the (very inexperienced) reviewer wanted to know what the menu for the award dinner was. For WP:DUE weight, you can always fall back to what the sources cover. If the sources only mention Hume's financial situation briefly, then Wikipedia should emulate the brevity of their attention to it.

Have you read WP:GACN? WhatamIdoing (talk) 10:11, 5 December 2014 (UTC)

I read it, but I'm not sure that the reviewer has. I wonder if the GAC on clear and concise could include your words above. My reviewer seemed worried that s/he could not tell whether s/he was reading a quote or a non-quote. Myrvin (talk) 13:44, 5 December 2014 (UTC)

Sorry, that was a stupid suggestion. I thought I was posting at WT:GAC rather than here. (Of course you've read this page if you're posting here!) It's difficult to get changes to GAC itself, especially changes that tend to make the list longer. Have you invited the reviewer to read this page? I think that all reviewers would benefit from it, but you might call me biased on that point, because I'm one of the primary authors.  ;-) WhatamIdoing (talk) 14:10, 5 December 2014 (UTC)

I did give a link here, but s/he just said s/he had "read GAC, thank you." You might find the new part-review edifying: Talk:David Hume/GA4. Myrvin (talk) 10:18, 6 December 2014 (UTC)

WhatamIdoing: I feel I have opened Pandora's box. Take a look at Talk:David Hume/GA4 and tell me which of these hundreds of demands are actually GAC demands. We now have two reviewers and two sets of hobby-horses.Myrvin (talk) 07:56, 10 December 2014 (UTC)

It looks like you've got a few good suggestions amongst an enormous amount of nitpicking (not all of which is unreasonable) and a remarkable amount of "I like ____ better than what you chose", where ____ is not in the criteria. Did you happen to request a more thorough review than average (for example, to prepare for FAC)? People do that sometimes, and it can result in this sort of "Spanish Inquisition" mode. Other times, it's just a sign of reviewer's inexperience or uncertainty. WhatamIdoing (talk) 23:39, 10 December 2014 (UTC)

Source of the article : Wikipedia