Wikipedia talk:When to cite

A few comments

1. In general, one point that can be overlooked is "is challenged or is likely to be challenged". The second is a judgement call - the first is not. So no matter how "common knowledge-y" something is, if editors want citations, they should be there.
2. A whole paragraph with a single cite can get confused as pieces are rearranged, reworked and so on. The reference is likely to get lost as the article is mercilessly edited - and if the reference isn't online, the reworking editors are unlikely to check it (let's be honest).
3. Ultimately, I don't see the problem with overcited articles - which is probably why I write articles like Scott Tremaine. Underciting --- if I can't confirm something with a few minutes on the google, it's probably best that it's cited. If I can ... it's probably best that it's cited anyways. But it's more important in the latter case. WilyD 15:53, 13 June 2007 (UTC)

To reply to your points:

1. The recent discussion that led me to move this page was that "is challenged" can apply to any part of the article, and people are stupid enough to do exactly that - challenge common knowledge, or things that we simply don't cite in practice. So, to be blunt, no, we should not be citing common knowledge.
2. The "distance" between sequential citations to the same source is a judgement call. It should be done in such a way that it is clear from context where the information comes from. By the same token, people must exercise caution when rearranging cited material so as not to break that coherence.
3. "Ultimately, I don't see the problem with overcited articles" - your opinion does not reflect practice. Raul654 16:03, 13 June 2007 (UTC)

1. What constitutes "common knowledge" is very dependant on your culture, nationality, profession, whatever. People do challenge common knowledge for inappropriate reasons, but almost anything is not common knowledge to a large enough segment of the population. Any editor with more than a few edits will have seen good faith challenges to information they consider "common knowledge" - citing is the easiest way to avoid the kind of edit wars we get over things like whether North America is a continent.
2. I agree with these principles, I don't think they're much different from what I said, except that I assumed less dilligence on editors tracking citations as they edit.
3. This only indicates that the featured article process is having problems, not WP:V or WP:RS, which are much more reflective of widespread opinion. If anything, it probably just means more eyes are needed there. "Overcited" is a patently absurd complaint. WilyD 17:12, 13 June 2007 (UTC)

Surely you would admit that this article is excessively cited. - Merzbow 17:45, 13 June 2007 (UTC)

The short answer is : No. Citations are the key to reliability, and eventual "widespread acceptance as reliable". Unless the number of citations made something "fundamentally unreadable", I wouldn't call it a problem. Cites not only add credibility and authority, the provide guides to further specific reading (and as an editor, direct me back to sources for more info to expand articles with). WilyD 19:48, 13 June 2007 (UTC)

I don't think there is anything obviously wrong in this proposal, but it is woefully incomplete. No mention is made of the host of situations where it may not be necessary to provide a cite... but where giving a citation is a good idea. Perhaps a "When a citation is recommended" section should be added to handle the grey zones. As WilyD says... it is better to overcite than undercite. I sum the whole issue up as "when even remotely in doubt as to whether a cite is needed... provide one".

" No mention is made of the host of situations where it may not be necessary to provide a cite" - Did you actually read the proposal? The cryptically named "Situations where it is not necessary to cite a source" takes up half the length. Raul654 16:05, 13 June 2007 (UTC)

Raul, the key part of my comment is in the second half of the statement: "...but where giving a citation is a good idea". There are tons of situations where one should give a citation, even if it isn't needed. Blueboar 18:38, 13 June 2007 (UTC)

Middle ground needed

This page curently outlines a list of things that must be cited, and another list of things that clearly do not require a citatiuon, and at least by implication should not be cited even if challanged. It does not discuss at all stuff in between, where a cition may be desireable but not mandatory. Obviously thsoe are largely going to be judgement calls, but soem guideliens might help inm making those judgemets. And soem examples of the kinds of things that fall into that broad middle ground should be provided, to help editors understand where thet middle ground falls, IMO. Also some discuion or guidence on how far apart citations should be might be a good idea. Also, some discussion of the issues of what sources to cite when there are multiple availabe sources might be a good idea. DES ^(talk) 16:21, 13 June 2007 (UTC)

It does not discuss at all stuff in between, where a cition may be desireable but not mandatory... Some examples of the kinds of things that fall into that broad middle ground should be provided" - such as? I'd be happy to write a description how of we do things in practice, if someone can describe specific situations that are not covered by this page as it currently exists. Raul654 16:26, 13 June 2007 (UTC)

On one hand, I feel that it is a sad thing to see that this page has been written (although I fully understand why); one the other, it's a good thing because perhaps this will avoid a few edit wars. That being said, I have two observations to make here:

• When in doubt, a person should consult one of the standard guides of style for professional researchers that discusses this issue: The Chicago Manual of Style, the MLA Style Manual Turabian. Citing your high school English teacher (or a similar unpublished authority) is not suitable because those people are not verifiable.
• When to cite is one of those situations where, as DESiegel points out, usage is not clear-cut; to borrow the terminology from the Internet RFCs, there are situations where citations may be used, situations where citations should be used, & situations where citations must be used. This gradation is important because it appears some Wikipedians would rather think in terms of absolutes, & tag everything without a source unless told otherwise. "Must" situations are those where the fact is cited or removed; "should" are those situations where a simple {{fact}} is enough; & "may" are those where an editor could add a tag, but only a tendentious editor would insist that it is needed.

I'll try to come up with some examples to illustrate the "may" & "should" cases. -- llywrch 18:32, 13 June 2007 (UTC)

My examples, based on Roman history:--

• The statement "Augustus was one of the greatest emperors of the Roman Empire" may need to have a citation. If the article supplies sufficient details of his reign, then finding a cite for this statement (& not a good quotation that says the same thing), IMHO, is overkill; the fact is obvious.
• The statement "Trajan was one of the greatest emperors of the Roman Empire" should have a citation. I could see if an article was written with enough information, this statement would be a reasonable conclusion. However, I'd feel more comfortable if this statement was supplied with a citation in any case: here there one can imagine a number of opposing POVs, and the evidence is not as familiar.

One other consideration: because this version of Wikipedia is written in English, like it or not there will be a systemic bias here. Our readers will be far more familiar with Britney Spears than the equivalent pop star in Russia or China; our preconceptions over what needs to be cited will reflect this. -- llywrch 19:19, 13 June 2007 (UTC)

Summary style

The quoted text no longer matches the current recommendation in Summary style. The text was changed recently by SlimVirgin. Effectively, the old text was arguing the same point about lead-section vs body applies also to summary-article-section vs subtopic. I'm quite happy with the lead section argument because the reader can find the references in the same article. In addition, one can review (for the purposes of GA/FA) the article on its own. I support the change by SV. I don't like the idea of one WP article relying on references in another article. Have there been any FAs where a whole section was allowed to be largely unsourced because the refs were in the subtopic? Colin°^Talk 17:22, 13 June 2007 (UTC)

I mostly agree. In an ideal world, all sub-articles of an article would be of the same quality and stay that way, and thus, there would be no problem. In the real world, an article like Islam can achieve FA while its subarticles remain ill-referenced and badly-written. Therefore I think it makes more sense that an article reference its own statements. - Merzbow 17:37, 13 June 2007 (UTC)

Agree with Colin; summary style doesn't exempt from citation. SandyGeorgia (Talk) 23:23, 13 June 2007 (UTC)

What to cite

(A) Is it relevant to mention WHAT to cite? For example, I've had people cite an entire book, without even giving a page number, claiming it is adequate. (B) I've also have people cite a source claiming that what it does NOT mention backs-up their statements. e.g. The study of UFOs is not studied by scientists [Source: Academic books (that do not mention UFOs). --First Base 19:35, 13 June 2007 (UTC)

Usefulness

I don't see how this article can be useful across the range of topics on Wikipedia. For example, citation needs, customs, and standards for medical articles are dramatically different than in many other topic areas, and I don't see how a list of what to cite can cover everything that needs to be cited in a typical medical article. Likely to be challenges is too vague, but trying to make a list of what to cite will be bound to miss a lot. SandyGeorgia (Talk) 23:27, 13 June 2007 (UTC)

It will get better over time and with input from more users. It's a big step in the right direction from what we currently have, and to be blunt, I'll take "not quite perfect" over nonexistant any day. Raul654 14:38, 14 June 2007 (UTC)

Well, the idea that we can summarize what needs to be cited in a medicine-related article scares me; the short answer is that almost everything needs to be cited in anything related to medicine, and I don't really know how to address examples in this format. Perhaps someone else knows how to get a start on it, but the idea of uncited statements in the realm of medicine makes me uneasy. SandyGeorgia (Talk) 18:21, 15 June 2007 (UTC)

It appears to me that some standardization of practice here would be worthwhile, so we should give it a shot and see if consensus can be reached. As an FA newbie the one thing that gave me the most heartburn was worries about referencing, given the number of extremely contentious threads on this issue that I've seen. Certainly any citation standards specific to narrow subjects like medicine and modern music are best handled via Wikiproject-specific guidelines (like Wikipedia:WikiProject_Music/MUSTARD), but I think there is much commonality. - Merzbow 04:48, 14 June 2007 (UTC)

(edit conflict) Can I offer my encouragement in your efforts to create a guideline on this. I have witnessed quite a few edit wars and squabbles over citations in articles, with people forming camps ranging from the 'citations get in the way and should be avoided if at all possible' to 'even general knowledge needs citing' and every possible place in between. (Check out Talk:Medieval cuisine if you want an example of a debate which keeps flaring up). It would be good to have something to indicate Wikipedia-wide consensus. Gwinva 14:12, 14 June 2007 (UTC)

Talk:medieval cuisine is a perfect example of polarized debates where one side is relying entirely on extremely partial interpretations of policy to get their pointless arguments across. There's no logic logic or common sense in all those random "THIS ARTICLE SUCKS BECAUSE I CAN'T COUNT ENOUGH FOOTNOTES"-claims. I'd rather put up with random bouts of whining than to honor drive-by fact-taggings by people who aren't going to verify those sources in the first place.

Peter ^Isotalo 15:05, 14 June 2007 (UTC)

I think that this is a really important concept to work on and I think it is one of the sources of confusion with the No Original Research issue. Wiki has tonnes of articles on readily accessible things where people are applying WikiLawyering approach to citations. It should be perfectly reasonable to take a book or product that is generally available and invent a description of it because any reasonable person can look for themselves - it is not research it is observation. So different styles of articles need different approaches in part because of the accessibility of the source data. History - historians have better access to source material than the public, so will tend to have more verifiable information, science, there are specialists, Harry Potter, its out there on a plate for all to see. Spenny 19:12, 14 June 2007 (UTC)

Will this conflict with WP:REF?

I note that a lot of this is already dealt with at WP:REF#When_to_cite_sources. This worries me. One thing that you will need to do if this proposal is ever going to be accepted is to make sure that you are not creating a conflict between two similar guidelines (I have seen this happen... for example: those working on WP:RS making edits that end up conflicting with WP:V). I am not saying that there is a conflict as the proposal stands now, just that the potential for it to do so is there. That potential exists whenever you have policies and guidelines that deal with similar topics. Please make sure you contact the folks who work on WP:REF to make sure that cross policy consensus is reached. You will also have to check the WP:REF guideline frequently to make sure that some future edit does not conflict.

One thing that is different between that guideline and this proposal is that this proposal attempts to answer the question: "when DON'T you need to cite?" I am not sure I agree with everything the proposal currently says on the subject, but I do like the attempt to answer the question. Blueboar 14:05, 14 June 2007 (UTC)

The problem, as is stated in the first sentence, is that "All material that is challenged or likely to be challenged needs a source." is a terrible standard. It is hopelessly vague, and doesn't address the obvious (and recurrent) problem of 'what happens when people demand citations for everything?' Raul654 14:13, 14 June 2007 (UTC)

Understood... that's why I think this is a good proposal, even if it ends up not being accepted. It is a question that needs to be asked. I just wanted people to be aware that there is a guideline that deals with this (however poorly) in existance, and to express my concern about the potential for conflict. The creation of or changes to one guideline have impact on others.

I agree with you that some citation requests are just silly, and should be ignored ... But we probably disagree about where to draw the line (ie I draw the line much closer to "cite everything" than it seems you do). Blueboar 14:29, 14 June 2007 (UTC)

I think the eventual hope is that this article would become the main article for WP:REF#When_to_cite_sources, which would be a summary. - Merzbow 17:58, 14 June 2007 (UTC)

Yes, exactly. Raul654 17:59, 14 June 2007 (UTC)

Can you give an example of challenged material that may not require a source? As far as I can see, "common knowledge" may fall into this category, though sources will easily be available. --First Base 21:18, 14 June 2007 (UTC)

I have seen real-life, on-wiki instances of people requesting citations for every scenario listed in the "Situations where it is not necessary to cite a source" section. Raul654 21:29, 14 June 2007 (UTC)

Apply to only above B-Class

It takes a lot of effort to find and add an in-line citation, but little to no effort to remove it. Given that most articles below B-class and some B-class articles lack citations, I don't think we should have a policy/guideline that gives examples of situations where it is not necessary to cite a source for B, Start, and Stub-Class articles. In particular, we should encourage (or at least not discourage) that every sentence in B, Start, and Stub-Class articles be cited (except for the lead paragraph). I think that use of this policy/guideline should be for those who decide the GA, A, and FA class rating and to assist them in resolving their citation disputes. And yes, it is something we need at the GA, A, and FA class level (perhaps as part of their criteria), so kudos to your efforts, Raul and Merzbow. -- Jreferee ^(Talk) 22:01, 14 June 2007 (UTC)

I find that an eminently reasonable suggestion. Raul654 22:02, 14 June 2007 (UTC)

"Challenging information" section

If I'm guessing Raul's intentions correctly, this section is intended to contain guidelines for challenging material in an article and for responding to those challenges. Would this be a good start:

• Any editor has the right to query why a statement is uncited. Those editors who believe the statement should remain uncited must respond in a timely matter, and indicate which category in the "Situations where it is not necessary…" section they think the statement falls under. If no response is forthcoming, the editor may remove or fact-tag the statement.
• If a response is made, and the editor who made the query disagrees, then that editor can challenge the statement. A challenge must be an informed argument, not a simple demand for a citation. A challenger can argue why he thinks a statement falls under a particular category in the "Situations where it is necessary…" section, or why he doesn't think it falls under the claimed category in the "Situations where it is not necessary…" section.

We should probably also add guidelines about how to challenge/respond to a challenge against cited statements. - Merzbow 22:54, 14 June 2007 (UTC)

My specific intent with that section was to stricly prohibit fact bombing (case and point) If people want to use that section for additional guidelines, that works for me too. Raul654 03:12, 15 June 2007 (UTC)

Heh, amusing diff. I find Jreferee's suggestion above relevant; an article that's at GA class and above has already passed a certain bar of quality, and thus the burden of proof should be on challengers to demonstrate why something should change. - Merzbow 03:28, 15 June 2007 (UTC)

Proposal motivation not clear

It might help to have a clearer idea behind the motivation for this proposal placed at the top of the proposal. Why do we need to clarify the Featured Article Criteria and Wikipedia:Verifiability statements and their interpretation? Really, what is wrong with requiring every sentence in a Featured Article be cited? As stated at Wikipedia:Featured article criteria, a featured article exemplifies our very best work and features professional standards of writing and presentation. I'm guessing the battles over this issue relates to attribution vs. professional presentation/professional standards at the Featured Article level. A statement such as "1 + 1 = 2" in Featured Article can be attributed with a footnote, but is it really a professional presentation or even professional writing to do so? The "sequential facts taken from a single source" portion of the proposal seems more of a proposed professional presentation for Wikipedia Featured Articles. However, the above comments seem to focus on attribution concerns. I think the proposal needs some clarification that it is meant to balance Wikipedia need to present its Featured Articles in a professional manner with Wikipedia's attribution policy (if that is in fact what the motivation is). -- Jreferee ^(Talk) 03:23, 15 June 2007 (UTC)

Your first question ( Why do we need to clarify the Featured Article Criteria and Wikipedia:Verifiability statements and their interpretation?) is already addressed on this page - Both of these are hopelessly vague, and have led to a number of disagreements in their interpretation. If anything, this is a great understatement.

Your second question (Really, what is wrong with requiring every sentence in a Featured Article be cited?) is because doing so produces articles that, frankly, look idiotic. (And, as I noted above on this talk page, are challeneged at FAC on that basis and don't pass) Raul654 17:13, 15 June 2007 (UTC)

As to your last comment ("I think the proposal needs some clarification that it is meant to balance Wikipedia need to present its Featured Articles in a professional manner with Wikipedia's attribution policy") - as I said above, I find this a reasonable thing to do, although I wouldn't limit it strictly to featured article. I would say it's designed to balance presentation versus attribution in all high quality articles, including B, GA, and A class articles. Raul654 17:16, 15 June 2007 (UTC)

Common knowledge

At the risk of a death spiral of definition, I think the concept of common knowledge needs discussion. Elsewhere I noted the idea that the test is that of what a reasonable person might know. The reasonable person test runs throughout Wiki, and seems to be one of the pitfalls that Wiki is battling with. For example, anyone can challenge whether something is known based on their own lack of knowledge, but there has to be a (vague) test to avoid the spurious. One example I came across (which I won't cite as I know the other person was acting in good faith) simply didn't have the common knowledge that you would have if you were of a certain generation, with certain interests. It was interesting how the edit war evolved, where I could point to a Wikipedia article which was entirely consistent with my statements, yet I was citation trumped with a dictionary definition (which did not understand modern usage). So common knowledge also has to be mindful of the specialisation.

I guess the question is how big does the pool of knowledge need to be, and in what context? Statements such as Napoleon being a French leader really should not need citation - it makes Wiki look stupid, I'd be quite happy with a statement about Winston Churchill being a great wartime leader, I'd be quite happy with comments about George Bush being derided as a leader by some people because it is so widespread, even if the supporters would be irritated by the comment and seek to censor it via citation lawyering. In the end that is what we want to stop, citation lawyering.

As has been coming out elsewhere, one of the objections to cite everything is that it makes Wikipedia look stupid. If the result of Wikipedia is that articles are full of proofs that the sky is generally blue it sends a message that the editors have little confidence in their own knowledge.

So there always has to be some consensus. Remember the purpose of avoiding citation nonsense is in part to protect evolving articles and newbies. People here are thinking in terms of citation as verifiability whereas I am concerned about the citation brigade who use it as a means to destroy articles where they disagree with content. Spenny 08:08, 15 June 2007 (UTC)

I note that this conversation is fragmented... with some people talking at WP:ATT (where it seems to be focused on what constitutes a legitimate "challenge" and what does not), some people talking here (focused more on when you don't have to cite) and some (I am sure) at other places. May I suggest that we pick one common place to discuss the issue, and let everyone know to discuss it at that place... so we don't end up with differing senses of consensus and decisions reached. Blueboar 15:03, 15 June 2007 (UTC)

Added section for grey zone situations

given some of the comments above... I think there is agreement that there are some situations that fall into a grey zone... where a citation should be given, even if one is not required. so... I have added a section for these kind of situations. It is far from perfect, and far from complete ... so feel free to rewrite or add to it. Blueboar 17:33, 15 June 2007 (UTC)

I don't particularly care for this section for two reasons - first, the other situations listed here have been quite specific (quotations, lead paragraphs, Sequential facts taken from a single source, 'etc). This new one - "contentions articles" - is not very helpful. Furthermore, the guideline itself - "You may want to use more citations on these" is not very helpful. While I don't disagree with the premise itself (that contentious articles could use more citations), but I'm very strongly leaning towards deleting that section unless it can be greatly refined from how it currently stands. Raul654 17:36, 15 June 2007 (UTC)

I also disagree with the section as written. The "Opinion" category above already makes it clear that statements disputed by reliable sources should be cited. Demanding citations in an article for undisputed statements is disruptive editing, no matter how controversial a subject. This is not to preclude editors being responsive on talk, which I think can be expanded upon in the "Challenging information" section. - Merzbow 17:50, 15 June 2007 (UTC)

Don't take it as written... It was meant as a start. The point is that this issue is not always black and white, yes or no, "you have to site this, but not this". There are situations where the same statement can go uncited in one article, but not in another - for example, a statement such as "Catholic's are not supposed to have an abortion"... you probably do not need a citation for this in an article on Catholic Dogma, but you probably should give one in the article on Abortion. (perhaps a bad example as this statement should be presented as an opinion and attributed as "according to the Church" ... but I hope you see my gist).

Some discussion of situations where a citation should be given (as opposed to must be given) should be included in the proposal. Even if you don't give an example... you need to acknowledge that such situations exist. Blueboar 18:51, 15 June 2007 (UTC)

Your "Catholic" example is covered under "Subject-specific common knowledge". When "Abortion" is the subject, that fact is not subject-specific common knowledge, so it should be cited, but if the subject is "Catholic Dogma", then it is subject-specific common knowledge, so it should not be cited. Note that this proposal is intended to be a guideline, not an absolute. A guideline is by definition something that isn't required to be applied as written, only suggested (albeit strongly). - Merzbow 19:49, 15 June 2007 (UTC)

Relationship to the policy

Is there a danger in creating a page to "clarify and interpret" policy? The whole point of the policy wording — that "material that is challenged or likely to be challenged" needs a source — is that it's deliberately vague so as to allow ample room for common sense. Any further pinning down may open loopholes for edit warriors. SlimVirgin ^(talk) 18:23, 15 June 2007 (UTC)

The problem is that there have been innumerable fights over subjective interpretations of policy - *especially* on featured article review, and to a lesser extent on featured article candidate. (I'm no expert on GA, but I suspect it's happening there, as well as on innumerable talk pages on Wikipedia). People come in and demand citations for everything, {{fact}} bombs the article, etc. Another person reverts. Unpleasantness ensues. The goal of this page is to reduce this. Raul654 18:28, 15 June 2007 (UTC)

If that's the goal, I fear the page may only open broader and more bitter disputes, whereas current citation differences are generally confined to few editors. But, time will tell. SandyGeorgia (Talk) 18:31, 15 June 2007 (UTC)

Example of citation density in a science article

I think we should aim to reproduce the same level of density of citations in Wikipedia as in the best-quality review articles seen in any particular subject. An example of citation in a biochemistry review is seen here Review article - Ubiquitin-binding domains. Here the general rule is that if you make a statement based on somebody's work, you cite their publication. In practice, this means most statements of fact are backed by a citation. TimVickers 18:55, 15 June 2007 (UTC)

Yeah I think it's a good criteria for wikipedia. Also, in case of multiple possible references for a statement, the "best sources" should be used; I think this should be an explicit requirement for Featured Articles.--BMF81 19:16, 15 June 2007 (UTC)

BMF81, I like that suggestion. Maybe it can be worded something like "If more than one source is available to support a statement, only the highest-quality source should be cited"? The exception of course is when one is trying to demonstrate specifically that a given opinion is widely-held, like "The album was positively received by critics". - Merzbow 19:52, 15 June 2007 (UTC)