Wikipedia talk:Citation overkill

Proposal

I would like to propose that we make this a policy since it is being overused on pages all over the project. All too often, people believe that notabilty might not be established with three citations, and they add a few more just in case. Unfortunately, this ends up loading the page up. I believe that establishing this as a policy will help to shrink article sizes as well as removing the obnoxious overuse of citations on articles. Kevin Rutherford (talk) 17:37, 9 December 2009 (UTC)

No, let's such leave it as the informal essay it is. Fences&Windows 22:52, 29 January 2010 (UTC)

Policy or guideline, seems some support exists at the village pump for this.Camelbinky (talk) 01:54, 31 January 2010 (UTC)

I think it's better as an essay. It's reasonable advice, but not suitable to the bright line response that policies/guidelines tend to evoke. It's adequate to say "a million cites doesn't improve an article" along with the reasoning and let the reasoning carry its own weight.--Father Goose (talk) 05:38, 4 March 2010 (UTC)

Counterargument

There are at least two occasions when "overkill" is appropriate. First, if an article says something like "many reports say the building was painted white", then it's nice to see many citations. Second, when a truly extraordinary claim is being made it's worthwhile to have many sources to support it. For example, an article I brought to FA covered a claim that the Astrodome would fly. As a reader, I think that's a assertion that requires extensive and impeccable sourcing. One thing this essay doesn't mention is that multiple citations can be merged into a single footnote. Doing so make it easier on the reader.   Will Beback  talk  13:56, 1 February 2010 (UTC)

In my drastic edit, I think I addressed that type of situation <g>. Collect (talk) 14:10, 1 February 2010 (UTC)

The only reason that this exists is because there is a, "Gee, I think a few citations is enough..." attitude among some readers. Too many of something can be a bad thing, and over three citations is just taking up space on the page. Kevin Rutherford (talk) 05:06, 10 February 2010 (UTC)

Oppose in its entirety

I oppose this in its entirety. All articles short of FA status are woefully under- or (not-at-all-) referenced as it is, and giving anyone ammunition to say that lack of referencing is A-OK is a really bad idea. → ROUX ₪ 17:21, 3 February 2010 (UTC)

For "fully referenced" - look at Astor House, Shanghai and tell me who will ever actually go through all its references.

For "reference overkill" see any number of political articles. MAny times there are ten or more references for a single word.

If three references are insufficient to back up a statement, chances are the statement should be reworded. Collect (talk) 18:07, 3 February 2010 (UTC)

Doesn't matter whether anyone will go through them, it matters that anyone can. Political articles are nasty POV battlegrounds. More referencing is good, because it guarantees to our readers that the information is not only accurate, it is overwhelmingly so. We are an encyclopedia; flow of reading style is a secondary consideration to providing accurate information that is reliably sourced. → ROUX ₪ 18:28, 3 February 2010 (UTC)

Wow, that article needs a lot of work with combining the references. I think that allowing for a crapload of citations on something as simple as the person not being able to die their shoes at the age of 13 can encourage more dumping of citations onto the fact. Adding more citations to a pretty obvious fact could potentially swamp the article in useless text that makes editing the real stuff harder because you have to sift through tons of crap to find the real point. Kevin Rutherford (talk) 05:17, 10 February 2010 (UTC)

To the point -- does seeing a word with 13 cites after it actually improve the readability fo an article? Ever? Does anyone ever actually read the references? (Empirically, many references are marginally connected to the topic of the article, at best). Collect (talk) 12:19, 10 February 2010 (UTC)

That and they rarely will read any reference unless the fact interests them. From what I have found, the references are usually parrots of each other, with most mimicing what the other is saying. Kevin Rutherford (talk) 18:31, 10 February 2010 (UTC)

IMHO such parroting is used to create an impression of overwhelming support for a point of little intrinsic value. Less is more - what is the point of having more than 3 citations in support of any single fact? --TraceyR (talk) 23:15, 10 February 2010 (UTC)

That's why having this become a policy would be useful because it would let editors know where the line is on citation dumping. Kevin Rutherford (talk) 13:26, 11 February 2010 (UTC)

• I think that references are a good thing, I agree that readability is a concern but I think that we should focus on better tools for managing lists and groups of references rather than deleting them. At the end of the day, I think that Wikipedias greatest contribution is exactly the collection of references that pertain to a particular topic. Unomi (talk) 19:21, 8 April 2010 (UTC)

I don't think that anyone is arguing against references - of course they are crucial. But do we need more than e.g. three sources supporting a single fact? That is the issue, surely? --TraceyR (talk) 05:59, 9 April 2010 (UTC)

I didn't mean to imply that anyone was against references, just that I am very strongly for them :). In some scenarios it protects against 'weathering', link death, slight rewordings that make some refs obsolete etc. It also offers a number of sources which might be used for later expansion or rewrites as they are unlikely to be containing the exact same verbatim information. You may have introduced a fair few underused citations when the article was a stub for example, by shedding those refs it makes it harder for later editors. It also plays a factor on contested articles, where shifting opinion may hold sway regarding what constitutes an extraordinary claim etc. Unomi (talk) 10:43, 9 April 2010 (UTC)

Personally, I created this article because you will get a basic fact like, "The sky is blue," followed by about 20 citations. People either seem to want to cite the obvious or there is a pervasive fear that it will be challenged without a ton of citations. Things really can't be done except remove them as they could theoretically take up half the page. Kevin Rutherford (talk) 23:27, 12 April 2010 (UTC)

• FWIW I generally oppose the approach and issue identified in this essay. I'm not sure how I would approach editing it, given that I wouldn't so much want to improve it as completely change it. i agree with one element: that lots and lots of cites after a word, to help bolster the position for that word, impairs readability. But that is it. Sometimes those many cites are necessary—even desirable—to communicate to readers that the choice of word is significant; and to communicate to new editors that this word was not chosen lightly, so please don't go messing with the consensus without a serious discussion at the talk page. So multiple cites around contentious points and phrases can be useful for both readers and editors. On a separate issue, I support the use of cites attached to the relevant facts - regardless of whether that is mid-sentence or not. We should take a good long-term view about WP. Fifty years from now, we want the research of the articles to both stand up to scrutiny and be easily reproduced. I don't want to spend time borrowing a book from a library only to find it turned out not to be the cite for fact A, only for fact B, and then have to get some other book to check fact A. As far as possible, cites should be linked to facts. While multiple cites can be a minor impediment to readability, i'd be interested to see the evidence that single cites are a significant problem. But for me, those single cites in the middle of paras and even, occasionally, sentences, are important to the credibility of the underlying research. hamiltonstone (talk) 01:35, 7 June 2010 (UTC)

WP articles are not intended as "research" but statements of what is generally found in secondary sources. Therefore multiple cites (in some cases 20 or more for a single word) are disruptive to the project in esse. If one wishes to make a research paper, WP is not the best venue at all. Collect (talk) 11:25, 7 June 2010 (UTC)

You have confused "research" with "original research". WP articles do not contain original research, but they must be researched by their writers. hamiltonstone (talk) 22:19, 7 June 2010 (UTC)

Template

Is there a maintenance template that can convey this? -- œ^™ 15:02, 26 April 2010 (UTC)

Nevermind, found it: Template:Too many references. -- œ^™ 15:05, 26 April 2010 (UTC)

Proposal or essay?

This is plainly not an active proposal, but my correction of the template was reverted. If anyone does want to actively propose that this be made into a guideline or policy then please do, otherwise this should be marked as an essay. Fences&Windows 15:41, 25 May 2010 (UTC)

Note the comments above where this is being discussed. And also note WP:DEADLINE Collect (talk) 16:00, 25 May 2010 (UTC)

DEADLINE is irrelevant, this isn't an article. If you want to propose it actively as a guideline or policy, do so. If it's actually an essay, then we should mark it as such instead of cluttering up Category:Wikipedia proposals. Sorry for posting a comment on your user page, I'm a muppet. Fences&Windows 22:04, 27 May 2010 (UTC)

Perhaops the fact that most of the proposals which are in similar state have been around much longer than this one should be noted? I do not see you marking all of the really old ones for removal - perhaops you should examine them by age? Collect (talk) 10:15, 28 May 2010 (UTC)

I wrote it and support but I have obvious bias since I nominated it as a proposal. What do others think? Kevin Rutherford (talk) 01:15, 8 July 2010 (UTC)

• Support making into a style guideline. As in those examples given it adversely affects readbility and just looks silly/amateurish. This is something all editors should strive to follow project-wide. -- œ^™ 10:51, 11 July 2010 (UTC)

• Oppose moving this beyond essay status. As an essay, it provides a useful perspective -- I would say an exaggerated/overreaching perspective, but that's fine in an essay. As a guideline, many of its specific points are either contradictory or inappropriate (or at least have not established consensus of their appropriateness per previous discussion). I will lay out further reasoning below. Such a proposed guideline presents problems of POV pushing via wikilawyering which have been brought up on this talk page but not resolved.

More broadly, this essay needs a broader set of eyeballs for vetting and demonstration of consensus before anyone should consider promoting it. Removal of reliable sources is a sensitive subject and if carried out too incautiously, it's liable to face serious backlash. If the provisions on removing reliable sources are to become binding, that needs to be reflected at WP:V -- so that is the place to establish consensus, not here. Style and content guidelines can be adjusted afterwards. Baileypalblue (talk) 07:03, 7 August 2010 (UTC)

I would not support the removal of existing sources. I am in support of a style guideline that directs editors to, in the future, not add absurd amounts of citations after a sentence just to prove a point. Furthermore, there are different ways of displaying/organizing inline citations, like in groups, that offer alternatives to a distracting string of numbers after a sentence. Any existing examples of citation overkill don't have to have their reliable sources removed per se, just restyled. -- œ^™ 03:31, 14 August 2010 (UTC)

• Oppose. Citation clutter is less of a problem than insufficient referencing. --_{Piotr Konieczny aka Prokonsul Piotrus| talk} 02:21, 11 June 2011 (UTC)

Provisions of this essay that should be reconsidered

• The nutshell: "Less is more when it comes to footnotes" -- fundamentally wrong attitude towards sourcing and a serious overreach. Good expression of the underlying perspective behind the essay, but potentially contradictory with some of the body (thinking specifically of How to trim excessive citations which touches on situations in which less is not more). The most one can validly say is that there will eventually come a point at which more cites to a statement become counterproductive. As it stands, the nutshell will be misused not only by vandals and POV pushers, but by mistaken good faith editors.
• "Where material can be supported clearly by a single footnote, best practice is to stop there." Wrong. Multiple cites demonstrate consensus that a claim is correct and put those claims into context the article cannot provide on its own -- who is making this claim? How widespread is this claim? What did they actually say? Others (e.g. User:Will Beback, User:Unomi, User:Hamiltonstone) have given reasons for using multiple cites which I won't recapitulate.

Multiple cites protect against linkrot and the relative inaccessibility of some sources. What happens when one source is published in a highly reliable but obscure journal, another is published in a source that is easily accessible online but relatively less reliable, and another is widely published but buried in hundreds of pages of technical jargon? Note that How to trim excessive citations touches on this problem by implicitly endorsing the use of multiple sources, which contradicts the assertion here that any time a claim can be supported "clearly" by one source "best practice is to stop there".

• "The next ten footnotes for the same point do not help anyone." The essay attacks extreme cases ("fifteen or more footnotes after a single word", Sixteen citations!, 17 citations for one sentence) in order to build consensus for opposition of more moderate cases. This is poor argumentation. What exactly is the intended scope of this proposed policy? Based on How to trim excessive citations it appears the objective is to establish a limit of three cites per point of information. If so, then references to "one footnote is enough" should be removed from the essay, and it should be written to convince us that four or five cites is overkill, instead of convincing us that fifteen cites is overkill. Of course it would be much easier to build consensus against fifteen cites than against five, so perhaps the proposed guideline should be scaled back to deal only with these extreme cases. If the ultimate answer is that we can't make the proposed guideline much more specific than it already is, that's a good sign it is too vague to be a guideline.

• "A rule of thumb is that one footnote after a sentence is enough, two begins to look untidy, and more than three is definitely clutter." Disagree. Fifteen sources = clutter, but four? Two are "untidy"?

• How to trim excessive citations: The details of this section are likely to engender controversy in application, even create an avenue for POV warring by selective deletion of sources one side in a debate dislikes.

• "The purpose of any article is first and foremost to be readable." This is an overreach. Readability is important, but verifiability is more important.

The stylistic concerns addressed by this essay are fairly trivial: a bunch of tiny blue numbers at the end of a sentence does little to impair readability for users, and any readability concerns for editors are secondary in an encyclopedia like this one which is intended to be optimized for readers. There should be a strong presumption in favor of preservation of information over issues of style, and this essay fails to find the proper balance. Baileypalblue (talk) 07:45, 7 August 2010 (UTC)

Those are some good points, and I will fix this up tomorrow if I can remember. If I don't do it in a few days, feel free to bug me as I won't mind a bit of a reminder. Kevin Rutherford (talk) 02:54, 22 August 2010 (UTC)

Okay, all done. Kevin Rutherford (talk) 00:22, 10 October 2010 (UTC)

How many, how few? A good question

The fact there is much discussion here and unclarity, flags it up as a good question. It seems like a tricky balance to strike - enough, solid citations on the one hand and cites not being used to bolster polemic or paranoia on the other. As quality of content is the main issue facing the public perception of WP it'll not be surprising if articles veer towards wanting a cite for everything posited as a fact, especially in BLPs. It seems articles were conceived to be written with the underpinning assumption of the good faith of the author. I wonder if this still holds, as WP moves towards emphasising demonstrable quality over good faith. I'm sure all this has been argued over countless times elsewhere, but the quandary become clear in essays like this. You'd have to find an answer to it before adding anything to MOS, I suggest. Spanglej (talk) 20:19, 8 September 2010 (UTC)

I should hope it still holds! Quality means brilliant prose, not verifiability. As we stop trusting our content editors to write good articles extreme deletionism starts growing out of control and we'll have idiots wanting a cite for the sky being blue. Only "material challenged or likely to be challenged" (WP:V) should have a cite and that's they way it should always be. -- œ^™ 08:57, 14 September 2010 (UTC)

Just another tool...

I fully understand the reasoning behind this proposal. And I fully concur that an overly cited passage can be aesthetically unpleasing. Yet, what concerns me is that if a finite number of cites becomes policy, it's just another tool for agenda driven editors to work their magic.

Let's say an article on your favorite topic contains a fact that you dislike. And this inconvenient fact is quite well cited. What you then do is use the "over citation" rule to prune some of the citations. But instead of pruning the least reliable on dodgy sources, you remove the most credible sources. Then some time later, you or another like-minded editor, comes along and goes "Whoa all these sources are garbage... let me go ahead and delete that for you." To me it seems like a policy that would be ripe for abuse. --192.35.35.34 (talk) 01:25, 16 November 2010 (UTC)

That's good point, but I believe it's being proposed as a guideline, rather than a set-in-stone policy.. and as for any policy or guideline, WP:IAR still holds. In fact, many of our other policies can also be abused by agenda-driven editors wanting to push their POV. We just need to be diligent in spotting these pointy and tendentious edits. -- œ^™ 02:08, 16 November 2010 (UTC)

This scenario assumes that no editors are going to notice that the better sources are being systematically deleted. I suspect that's not a reasonable assumption for most articles. The scenario requires (1) a POV pusher, (2) an excessive number of citations, (3) a significant variability in quality of citations, (4) nobody noticing the removal of the best sources [say, like the editors who added them in the first place], (5) nobody adding better sources, and (6) multiple edits spaced out over time [not something your average POV pusher does]. I'm not going to say that it can't happen or that it hasn't happened, but it probably affects far less than 1% of articles. WhatamIdoing (talk) 04:50, 16 November 2010 (UTC)

Template:Citation overkill

I'm proposing this citation overkill template. It probably shouldn't be used too often, but I'm lazy. :) Kayau Voting IS evil 13:08, 24 March 2011 (UTC)

Wouldn't adding another long inline string of characters next to the string of numbers just be adding to the problem? Maybe it would be better as a standard message box template instead? -- œ^™ 07:01, 26 March 2011 (UTC)

I'm going to convert it to the standard message box so it makes more sense. Besides, if you have to question this, you probably aren't counting the number of citations to the left of it anyways. Kevin Rutherford (talk) 16:48, 2 May 2011 (UTC)

Does no one see what a terrible idea this template is? Solve barely any problem, provides much potential for abuse. / edg ☺ ☭ 22:40, 2 May 2011 (UTC)

Um -- Wikipedia is well known for having single sords with dozens of superscripted cites following each one. I assure you that it does not make articles easier to read. Really. Collect (talk) 00:15, 3 May 2011 (UTC)

"Um", Wikipedia is "well known" for {{Citation needed}}. "Citation overkill" hardly merits a maintenance template. If you personally have trouble reading text with superscripted characters, you can add .Inline-Template {display:none} to your personal CSS styling file. / edg ☺ ☭ 11:14, 3 May 2011 (UTC)

What percentage of all WP users have a monobook file with that in it? Really? I would wager it is well under 1/1000000. And I would even give odds on that. And of non-editors, the ratio is likely even lower. So much for a useless suggestion for them. Collect (talk) 11:56, 3 May 2011 (UTC)

That few? Then maybe it's not as much of a problem as you think. / edg ☺ ☭ 14:52, 3 May 2011 (UTC)

All it does is annoy the hell out of many users. I guess that is unimportant compared with the value of 25 cites for the word "the" in an article. <g>. Collect (talk) 14:55, 3 May 2011 (UTC)

poorly constructed sentence

A well meaning editor may attempt to make a subject which does not meet Wikipedia's notability guidelines through quantity of sources ....needs fixing Penyulap _talk 21:56, 8 August 2011 (UTC)

Wikipedia:Overlink crisis

Wikipedia:Overlink crisis can probably be merged here. ---— Gadget850 (Ed) ^talk 10:29, 29 September 2011 (UTC)

William Evans (Medal of Honor recipient)

Is it really appropriate to include articles created by IPs though AFC? Afterall, articles without "in-line citations" are routinely declined on AFC because of this issue. It seems kind of unfair to force someone unfamiliar with wiki-formatting to go through the effort of converting their references into "wiki citations" and then mock their efforts. I realise this isn't the intent of the essay but I've rarely seen an anonymous user recieve help on AFC with regards to this issue aside from pointing to WP:INCITE. 129.10.104.177 (talk) 21:32, 3 November 2011 (UTC)

The problem doesn't cease to exist because of the source. This essay doesn't mock anyone, it outlines a problem and gives specific examples. New editors should not simply be excused for shortcomings in their efforts. Those issues should identified and guidance provided for making it better. --RadioFan (talk) 21:55, 3 November 2011 (UTC)

Nutshell wording

How about in the nutshell part of the lead we put the following:

"When verifying information it is better to have a couple of great sources than a stack of decent or sub-par sources."

Seems to be less like something that would give people a bad impression.--The Devil's Advocate (talk) 21:54, 26 November 2011 (UTC)

One thing I don't get

For those who are opposed to the argument of this essay, can't you just have one cite with multiple sources in it? E.g. "X then went to the Y to find Z."[1] followed in the reference list by "[1] Syett, 107. Garvus, 447. Meckler, 66. Munroe, 30. Alistus, 178." Then, of course, those sources are added in the Bibliography or References section of the article. Having [1][2][3][4][5][etc.] not only harms readability but also looks unprofessional, like the article is screaming, "Look at me! I'm well-referenced!" and trying without success to win the reader's confidence. --Mrdie (talk) 03:41, 16 December 2011 (UTC)

Same citation, different sentence

With respect to medicine

In technical matters it is useful if every sentence has a reference. Often sentences get moved around and the ref to that sentence than gets lost. I propose a better method than simply removing duplicate refs is to simply use <!-- --> such that when someone comes to add a tag they will see the ref that supports the text in question.--Doc James (talk · contribs · email) 08:32, 11 March 2012 (UTC)

Yes, articles do get edited, and when that happens, editors are responsible for making sure that the right citations are properly moved with the correctly corresponding passages. This is what I do. Besides, this applies only to when an entire passage is supported by one single source. If any portion of a paragraph or passage is not supported by that source, then that portion can indeed by addressed with a fact tag (if not outright removal). Nightscream (talk) 20:32, 11 March 2012 (UTC)

I agree that sentence-by-sentence duplication of citations is useful—for sloppy future editors. But we're providing citations for our current readers, not for the small proportion of sloppy editors in our midst who might, someday in the future, make a change to the article. The example at Wikipedia:Inline citation#Citation_density is a good example of when we don't want to repeat citations. WhatamIdoing (talk) 21:58, 11 March 2012 (UTC)

Using the no wiki tag makes it much easier for me to maintain articles and reduces confusion somewhat among our readers who are adding tags. I have had a number of sentences tagged in which the ref was only a single sentence away. Thus many of our readers it appears expect / wish citations on every line. Using the no wiki tag is a fair compromise IMO, simply removing refs is not. Doc James (talk · contribs · email) 04:39, 12 March 2012 (UTC)

I support what Doc James is trying to say, but to clarify, WP:NOWIKI is different from what he is describing, which is actually a WP:COMMENT or "invisible comment". I do think that the standard on Wikipedia should be that after every period or point everywhere except in the lead there should be a way to verify the source of the statement. It is extremely difficult to sort the source of references when only the last sentence of a paragraph has a citation which is supposed to cover the paragraph, but then the entire paragraph is somehow restructured. Wikis frequently change and a citation for every sentence is a preferable practice to alternatives of expecting continually vigilance or stifling new editors who have not yet learned all citation customs. Blue Rasberry (talk) 15:34, 27 September 2012 (UTC)

Same citation, different sentence (continued)

I disagree with this modification. Adding one citation for each sentence allows readers and editors to understand what is supported by a source and what isn't. It would be easy to slip in an unsupported sentence and put the citation at the end to force readers to assume that all of the information is supported. Alternatively, a reader may assume that none of the information is supported due to the lack of a citation. Ryan Vesey 02:37, 26 August 2012 (UTC)

The modification goes against the intent of this essay which was to keep things like "statement^{[1][2][3][4][5][6][7]}" from happening. Ryan Vesey 02:39, 26 August 2012 (UTC)

I don't think it "goes against the intent of this essay". There is a similar example given at WP:MINREF#Citation density, and sentence-by-sentence citations have never been required by any policy.

There are silly instances of supplying citations sentence by sentence, and there's nothing magical about sentence-by-sentence citations that prevent someone from slipping an unverifiable sentence in. Sometimes (especially with inexperienced editors) they even "steal" one of your superfluous citations for their new sentence, which makes it look like your sentence is unsupported and theirs is fully supported. WhatamIdoing (talk) 20:44, 28 August 2012 (UTC)

I never said they have been required by policy, but there's no reason to outlaw them in this essay. Ryan Vesey 20:52, 28 August 2012 (UTC)

It makes sense to put the same citation behind many sentences. Normally it wouldn't, but Wikipedia is constantly changing, and there are numerous editors who come and go. Information is constantly added to a paragraph, and when articles get rearranged and rewritten, its harder to keep intact. A citation behind a paragraph doesn't represent new additions to that paragraph. There is the problem of having not enough inline sources for large articles, and this is because initially a citation wasn't put behind every sentence to start with.

Think of making a fully referenced article into a good, higher class or featured article. Then there's the complaint, "not enough inline citations", when initially the article was well written and done without "overkill". The original author is gone. You have to go back and put citations back in, when there are over 100 redundant references with no inlines. It'd be easier to start that from scratch.

And if someone "steals" a citation to reword it. Wouldn't the citation back you up, or discredit the other user? A clear note of citations will help who has it right. The only way to prevent bad insertions is to have many watchers, and have every sentence cited. Leaving out citations discourages article improvement. Sidelight12 ^Talk 09:40, 13 December 2012 (UTC)

While a citation behind every sentence may not be needed, I do not see anything wrong with it and it is often a good idea. Thus I would disagree strongly with this essay. The issue we have know is that if a citation is NOT present behind every sentence someone comes and tags it with cn if a citation is behind every sentence some come and remove a bunch of them quoting this. Doc James (talk · contribs · email) (if I write on your page reply on mine) 03:29, 21 June 2013 (UTC)

Several unrelated ideas here and this page should be split

There are several ideas in this essay and they ought to be split because they do not relate to each other. Here are some of the ideas:

• Some information on Wikipedia, like assertions that the elephant is large, are simple and accepted statements which, while debatable by argumentative people, can be generally understood and accepted without a citation.
• There should be a limit to the number of citations listed after a sentence, because a long string of citation numbers is distracting.
• When an entire paragraph is sourced with a single citation, it is acceptable to place a citation only at the end of the paragraph and preferable to leave each individual sentence without a citation.

That third point about recommending that some sentences preferentially remain without citations is the one that bothers me and I do not like it connected to the other ideas. This idea makes statements in Wikipedia more difficult to verify as content is changed by other users and causes maintenance problems in other ways. I would favor a best practice recommendation of having a citation for every sentence or contained statement. If anyone ever feels strongly about this also then ping me and perhaps I would collaborate with others in drafting a proposal to make a recommendation - though not a hard guideline - that a citation for every sentence is best, and anything less than that is just acceptable and not preferred. Blue Rasberry (talk) 15:56, 24 January 2014 (UTC)

True. What would focus this essay is to consider the following example paragraph & the issues raised in properly providing adequate sources for the statements made in it:

These are the facts to be discussed in this paragraph, drawn from several sources, both primary and secondary. (1) The consensus view is this. [Alternatively, "Jones expresses the consensus view".] (2) However, Smith has proposed this interpretation (3) which is a notable minority viewpoint. (4)

(1) Whether to combine multiple sources into one footnote, or split them out so each source has its own footnote. This also leads to the issue whether comments on primary & secondary sources in the footnotes should be combined in one place, if multiple sources have been combined that way, or they could be split out to show the parts they relate to.

(2) If sources cited in (1) are repeated here, whether to separate them into their own footnotes to be repeated here & elsewhere in the article, or to combine them. And there is the issue whether we should consider it sufficient to provide one source to confirm the existence of a consensus amongst experts, or require more.

(3) Sometimes it is good practice to put footnotes only at the end of the relevant sentence, & sometimes footnotes need to appear in the middle of a sentence after the relevant clause for emphasis.

(4) Sometimes accepted practice is only to combine all citations into one footnote & list all sources there. This has the drawback that it can be difficult to match facts to the relevant source.

How & where to attach links to sources is often more of an art than an engineering problem -- something not always appreciated by those wishing to handle these problems. -- llywrch (talk) 18:22, 7 August 2014 (UTC)

36 citations

How about adding this article to the list? It has 36 footnotes at one place. (t) Josve05a (c) 23:04, 3 March 2014 (UTC)

Josve05a Go for it. Kevin Rutherford (talk) 04:47, 4 March 2014 (UTC)

WP:PAIC

NukeofEarl, hours ago, I saw this edit by you removing mention of WP:PAIC, and I didn't know whether or not your removal should stay intact or be reverted. Hours later, I've obviously decided to bring the matter here to the talk page. The reason that I'm confused by the aforementioned removal is because the sentence that was referring to WP:PAIC is about placing citations "at the end of the passage that they support." And WP:PAIC corresponds with that, stating, "The ref tags should immediately follow the text to which the footnote applies, with no intervening space." Flyer22 (talk) 19:28, 26 June 2014 (UTC)

Reverted, since I was ignored. Flyer22 (talk) 00:23, 29 June 2014 (UTC)

Tag

Isn't there a tag to put on top of an article with citation overkill? For example, the HDMI article regularly has three or four references for a stated fact and a total of 209 references, which seems like overkill to me. Especially since it's not at all controversial. I cleaned up a little, but i'd like to leave behind a tag to invite other editors to continue cleaning up.PizzaMan (♨♨) 23:46, 10 December 2014 (UTC)

Note: PizzaMan took this matter elsewhere; see Wikipedia:Village pump (policy)/Archive 117#Tag for too many citations. Flyer22 (talk) 10:31, 15 December 2014 (UTC)

In my opinion this is not something to complain about. There used to be tags saying "This article needs general cleanup" but those are now not allowed, because tags should be used to ask for help fixing a particular problem. When citations are overused, people might be tempted to remove them so that the tag can be removed, but I do not want anyone to be too quick to remove citations. Removing citations should be a thoughtful process that goes along with deeply editing the article, and not just one part of it. This is because it is a large investment to add citations, but much easier to remove them, and I would rather over-citation than undercitation.

Right now, I would like to not make a work queue which encourages people to quickly remove citations, and would rather than interested people just fix the problem when they deeply review whole articles or sections. Blue Rasberry (talk) 17:09, 18 December 2014 (UTC)

Changing kill reference

In the big scheme of things this is very low priority, but I was updating the Nice shooting article and realized after the fact that using citekill was potentially insensitive and inappropriate. What do you think about getting rid of citekill and citeoverkill as references and just using overcite and the less violent variations? I know it will be a hard thing to unlearn for some people, but in the long run perhaps it will be more sensitive. As far as I can tell all that has to happen is the two variations have to be removed from this article. Thoughts? ----

Citations

Our readers appear to want at least one citation supporting every single sentence for medical content.

When a single ref supports multiple sentences in a row I used to hide all but the last one. I however am going to stop doing this as this happens SO often [1]

It basically comes down to article maintenance. It is easier for me to maintain 5,000 articles when our readers are getting what they want.

I thus propose replacing

Extended content

In addition, as per WP:PAIC, citations should be placed at the end of the passage that they support. If one source alone supports consecutive sentences in the same paragraph, one citation of it at the end of the final sentence is sufficient. It is not necessary to include a citation for each individual consecutive sentence, as this is overkill. This does not apply to lists or tables, nor does it apply when multiple sources support different parts of a paragraph or passage. This is correct: In the first collected volume, Marder explains that his work is "about the affinity of life," wherein the characters "understand that ultimately they depend on each other for survival." Wiater and Bissette see this relationship as a wider metaphor for the interdependancy of the comics industry. Indeed, addressing the potential underlying complexity, Marder suggests that "it's harder to describe it than it is to read it." He also calls it "an ecological romance... a self-contained fairy tale about a group of beings who live in the center of their perfect world [and are] obsessed with maintaining its food chain," a self-described "really low concept!" Equally, he says, "the reader has to invest a certain amount of mental energy to follow the book," which includes "maps and a rather long glossary." Despite these potentially conflicting comments, Wiater and Bissette reiterate that "there is no simpler or more iconographic comic book in existence."<ref name="Rebels">[[Stanley Wiater|Wiater, Stanley]] & [[Stephen R. Bissette|Bissette, Stephen R.]] (ed.s) "Larry Marder Building Bridges" in '''''Comic Book Rebels''': Conversations with the Creators of the New Comics'' (Donald I. Fine, Inc. 1993) ISBN 1-55611-355-2 pp. 17–27</ref> This is also correct, but is overkill: In the first collected volume, Marder explains that his work is "about the affinity of life," wherein the characters "understand that ultimately they depend on each other for survival."<ref name="Rebels">[[Stanley Wiater|Wiater, Stanley]] & [[Stephen R. Bissette|Bissette, Stephen R.]] (ed.s) "Larry Marder Building Bridges" in '''''Comic Book Rebels''': Conversations with the Creators of the New Comics'' (Donald I. Fine, Inc. 1993) ISBN 1-55611-355-2 pp. 17–27</ref> Wiater and Bissette see this relationship as a wider metaphor for the interdependancy of the comics industry.<ref name="Rebels" /> Indeed, addressing the potential underlying complexity, Marder suggests that "it's harder to describe it than it is to read it."<ref name="Rebels" /> He also calls it "an ecological romance... a self-contained fairy tale about a group of beings who live in the center of their perfect world [and are] obsessed with maintaining its food chain," a self-described "really low concept!"<ref name="Rebels" /> Equally, he says, "the reader has to invest a certain amount of mental energy to follow the book," which includes "maps and a rather long glossary."<ref name="Rebels" /> Despite these potentially conflicting comments, Wiater and Bissette reiterate that "there is no simpler or more iconographic comic book in existence."<ref name="Rebels" /> One can hide citations with <!-- --> to prevent confusion in the future; then can be uncommented if new material with different sources is interpolated later. References can then occur after each sentence which is the preferred style for medical content.

with

"If multiple consecutive sentence are supported by the same reference one can:

1. have all references shown
2. hide citations with <!-- --> to prevent confusion in the future; this can be uncommented if new material with different sources is interpolated later.
3. put the reference at the end of the section but realize that ongoing maintenance will be required as other editors may mistakenly add {{cn}} tags or delete content as they think it is unreferenced"

Doc James (talk · contribs · email) 19:41, 6 May 2017 (UTC)

Doc, you want to remove all of that? Considering that it explains what citation overkill is and that citation overkill can be a pain and unencyclopedic, I don't support removing all of that. I am open to changing it, though. Also keep in mind that we have more than just medical articles to consider. I'll go ahead and alert a lot of WikiProjects and similar pages to this proposal; that will sort of serve as an RfC, but an official RfC may need to be started. Flyer22 Reborn (talk) 06:02, 9 May 2017 (UTC)

Relatedly, the current language "If one source alone supports consecutive sentences in the same paragraph, one citation of it at the end of the final sentence is sufficient" is in direct conflict with current DYK rules, which (for DYK-nominated articles) require the sentence or sentences containing the hook claim to have their own citation, even when it would be the same as the next citation in the same paragraph. —David Eppstein (talk) 06:30, 9 May 2017 (UTC)

• Oppose. I oppose the proposal. The current text explains things well, but the proposed text is unclear. Also, it's not a good thing to have the same ref repeated after every sentence. If there's a reason to do it on particular articles, that's fine, but it shouldn't be recommended as best practice. SarahSV ^(talk) 07:04, 9 May 2017 (UTC)

• I'm wondering three things. (1) The guideline is WP:CITE, so anything add to this essay that contradicts it won't have much or any force. (2) How many people add commented-out citations after each sentence? I've seen it at a few medical articles but otherwise not. If this is limited to medical articles, it would better to propose it at MEDMOS. And (3), James, you wrote: "Our readers appear to want at least one citation supporting every single sentence for medical content." Do you have a source for that? SarahSV ^(talk) 22:46, 12 May 2017 (UTC)

• I review a lot of edits and see a fair number of IPs removing sentences in the lead that are unreffed or have a hidden ref with the claim that it is not referenced. Doc James (talk · contribs · email) 22:52, 12 May 2017 (UTC)

• @Doc James: okay, thanks. The solution to that is to add the extra refs after sentences where you see it happening. I certainly think leads should be referenced, and WP:LEADCITE allows it. Another thing you can do is add an edit notice that says: "Sources are in the nearest footnote after the text in question", or words to that effect. SarahSV ^(talk) 23:09, 12 May 2017 (UTC)

• You wrote "The solution to that is to add the extra refs after sentences where you see it happening."
• I disagree. The solution it to add them before it happens. That's why we are having this discussion. There is no logical reason to wait for the problem to happen again. QuackGuru (talk) 23:50, 12 May 2017 (UTC)

I am a strong supporter of so-called "citation overkill" for two reasons: 1) this is Wikipedia and our readers cannot know for sure if editors have added unreferenced sentences in the middle of a paragraph unless every single sentence is referenced. 2) as an editor, whenever I want to expand a paragraph, I at least know which content is cited and which one is not. Finally, I don't see a problem at all with so-called "overkill". It does not make it harder to read at all. It just makes it safer!Zigzig20s (talk) 07:09, 9 May 2017 (UTC)

So you think having six or ten plus references for a single sentence is normally fine? What about sixteen? The essay allows some leeway (such as WP:CITEBUNDLE, which is what I sometimes use), but it discourages unnecessary and outlandish citing behavior. Flyer22 Reborn (talk) 07:20, 9 May 2017 (UTC)

No, sorry, that's not what I meant. I mean every single sentence should be cited, usually with one reference. We should stop at three references per sentence at the most, otherwise it's really too much. But unreferenced sentences are a nightmare, because there is simply no way to know if the content is made up or not (which is compounded by further expansions).Zigzig20s (talk) 07:28, 9 May 2017 (UTC)

Simply no way? The first step is to read the references mentioned in the same paragraph. The goal of verifiability is not to allow readers to skip the references - it is to allow those who want to read the references to do so. Only after looking at the references included in the article would someone have a sense of whether a particular fact is unsupported by the references. In any case, the idea that every sentence should be cited is far from mainstream on Wikipedia, as in real life. — Carl (CBM · talk) 10:42, 9 May 2017 (UTC)

Most readers do not have time to double-check references. As an editor, it also saves me time if I know for sure what is referenced and what is not. For example if there is an "early life" paragraph and I want to expand it with more content based on new references I've found, it's extremely useful for me to know if their place of birth, parents' names, and alma mater are all already referenced, or if some of this content isn't. Moreover, I do not see a problem with adding a reference at the end of each sentence because it does not make reading more difficult at all. I don't think "mainstream" is an argument here; we should focus on ways to improve the encyclopedia, not what some editors (on a hunch) are used to. In short, more references are a win-win situation for everyone.Zigzig20s (talk) 13:20, 9 May 2017 (UTC)

I agree with the approach outlined by Zigzig20s. We should aim to cite every sentence, especially for articles that might be more contentious, or those that are more likely to be expanded over time, such as BLPs. Articles get rearranged and sometimes paragraphs are combined or split into two. Thinking long-term, the sentence is the key unit, not the paragraph. If we also consider the increasing role of Wikidata in supporting and creating articles, then its role in supporting discrete facts (ie sentences), favours such an approach. One citation per sentence should be enough, unless something is particularly contentious, or perhaps where each source supports a different element of the sentence. Edwardx (talk) 13:34, 9 May 2017 (UTC)

Note: I alerted guideline and policy pages (like WP:Citing sources and WP:Verifiability) and WikiProject pages (the ones seen here) to the above proposal. I also see that the matter is being discussed at WP:Med; because it was already being discussed there, I didn't need to alert that project. It's noted there that this page is not a guideline. While that is true, it is commonly cited. I also think it should be a guideline. Flyer22 Reborn (talk) 07:20, 9 May 2017 (UTC)

I think putting a citation on every single sentence is silly. A reader should just look at the next citation if they are concerned about what is said. A citation per short paragraph is what I would consider the minimum and more should be put in only where necessary. That someone has put a citation on a sentence does not mean that the sentence really conforms to the citation. Checks on citations should be because someone thinks a citation does not support what is said, not because they have some rule about sentences.

This sort of rule will just lead to funny conventions in the source which people who put in 'citation needed' where unnecessary will not understand and it will be ignored.

For lazy editors who write their own thoughts and think any old citation supports it they will stick in this thing and people will have the erroneous idea that they have actually checked properly where they have not.

Overall I think it is a bad idea. The paragraph is the basic unit for conveying thought, not the sentence. If the sentences in a paragraph need individual citation then the paragraph is in essence a list of separate connected things rather than one thing. This may be quite often true in medicine but in most topics the whole of a paragraph is best treated as a single indivisible unit. Dmcq (talk) 08:38, 9 May 2017 (UTC)

Agree with User:Dmcq - this is exactly what I was thinking as I read above: a sentence is not the basic unit of exposition. Of course there may be cases (for example, laying out opposing views on something) where it really is sensible to have refs dotted through a paragraph, but it should absolutely not be a rule or guideline. I also find myself irritated on contentious issues to see strings of refs, as though this somehow added persuasiveness. I do not honestly think there should ever be more than 3 or 4. It should be sufficient to add other sources as general references. Imaginatorium (talk) 09:22, 9 May 2017 (UTC)

I agree. What are we going to do, cite the same source for every single sentence in a paragraph, if that paragraph summarizes what's in the source? Don't be ridiculous. Moreover, contrary to what the proposed wording seems to suggest, editors shouldn't be going around deleting material that is already cited according to this guideline, nor placing tags to compliant material. That's a problem with editors, not with the guidelines. Editors should be checking the sources, not mindlessly looking at the text to see if each symbol "." is followed by the symbol "^[". Sławomir Biały (talk) 10:38, 9 May 2017 (UTC)

Sławomir Biały, I think one issue is that the essay is not an official guideline, even though it's treated like one. Flyer22 Reborn (talk) 08:42, 12 May 2017 (UTC)

Could this be managed by software? I envisage something which allows all citations within a paragraph to be represented by a single symbol after the last full stop, regardless of how many are in the source text and how they are distributed within the paragraph. To get the actual picture, click on the symbol and all the citations in the paragraph are displayed in the usual way. Ideally this would be combined with linking each citation to the actual text it is claimed to support, which would be highlighted by hovering on the specific reference mark. Result would be clean, easy to read text with all useful citations linked to the relevant content. • • • Peter (Southwood) ^(talk): 13:11, 9 May 2017 (UTC)

This is an improvement, but ultimately this should be done programmatically or with a dedicated template, for instance {{span ref}}: Template:Span title ¿Or maybe we could take a cue from Spanish and put the refs at both ends of the series of claims they support? Bright☀ 14:10, 9 May 2017 (UTC)

References

I personally think we need readers to understand, as noted above, the sentence is not a discrete unit of information and not every sentence needs citation; if a statement is uncited immediately after the trailing period, one should scan until the next reference in that same paragraph and implicitly assume that is the source of information (with some common place exceptions). If we can't convince readers of that, then we do need a ref-spanning template/mechanism that a reader could mouse-over a superscript inline reference and the browser would highlight the text that that supported (which may be more than a sentence). However, this has all the potential to be broken inadvertently by editors since it would require a leading *and* trailing tag/markup. I'd rather go the education path, of making readers aware that not every sentence needs sourcing, which requires no changes to any article, just a few guideline pages. --MASEM (t) 16:00, 9 May 2017 (UTC)

Why can't editors simply cite every sentence? Much easier I think, especially for new articles.Zigzig20s (talk) 16:06, 9 May 2017 (UTC)

On second thought I agree with this, it's easier to explain to readers that they have to check the nearest reference than to handle leading and trailing syntax. Bright☀ 20:44, 9 May 2017 (UTC)

Comment. From the point of view of someone who only works on music articles, I'd say the correct way to handle citations in that text (under Extended content) is to repeat the Rebels reference whenever we give another writer's viewpoint. That's when a reader requires clarification, surely, not to mention other editors. Only the last of those sentences covering Marder would need it, for instance. (In total, I count 4 instances in the paragraph.) That's the approach I've been encouraged to adopt by reviewers, anyway. So I'd think the choices under "If multiple consecutive sentence are supported …" should allow for that. JG66 (talk) 17:13, 9 May 2017 (UTC)

I'm with Zigzig20s on this. I cite every factual statement, usually, because I want the reader to know what is said by what source & what's not. I get the "overcite" complaint, tho; I don't see a need for 3 or 4 separate cites for each sentence (or even paragraph): if the same cites are supporting the same assertion, "bundle" them under a single fn. IMO, that also makes it clearer for the reader: if 3 sources are supporting the same assertion, they shouldn't be listed as if they're for separate assertions. (Am I being clear, here?) My $0.02, anyhow. TREKphiler ^{any time you're ready, Uhura} 18:23, 9 May 2017 (UTC)

It isn't always desirable or possible to bundle; for example, if you're using {{sfn}}. This proposal seems to be suggesting:

This is a sentence. This is another sentence. This is a third sentence.

It would make articles harder to read and much harder to edit.

SarahSV ^(talk) 19:16, 9 May 2017 (UTC)

User:SlimVirgin, it is impossible or more difficult to verify a claim without a citation. It is much harder when admins continue to support editors who replace sourced text with original research on the articles I edit. A visible source at the end of the sentence could help slow down the admins. QuackGuru (talk) 01:53, 11 May 2017 (UTC)

Sarah: even given a need for multiple citations at a given point, there is no need to have a string of note-links (the ^[1][2][3]), nor multiple back-links (the ^{a b c}). In the first case, just drop the needed citations in a single note. E.g.: <ref>{{harvnb|...}}; {{harvnb|...}}; ... </ref>. Which is hard to do with {sfn}, as it automatically puts each citation into it's own note. So don't use it. Just drop the citations into a single note. In the second case: that arises "re-using" notes with named-refs (the <ref name=...> entities), which is implicit in using {sfn}.

Revising your example in this manner produces:

This is a sentence. This is another sentence. This is a third sentence.

There is some ambiguity in the nature of "Ref", but that is a separate matter. I find this "style" a lot easier to edit because I'm not having the hunt around the for the master named-ref. It's also shorter (by 50 some characters), and less cluttered. ~ J. Johnson (JJ) (talk) 18:02, 11 May 2017 (UTC)

• Support Doc James' suggestion and, indeed, multiple citations. Not having at least every sentence cited leads to, with new editors, stranding article text without a proper citation by interposition of new material (thus leading to claims that the stranded material is uncited). With experienced editors it discourages article improvement through fear of stranding or mis-citing material: I've had it happen that I wanted to add something to an article in the middle of a long stretch of material with just a citation at the end, but didn't have access to the cited material so didn't know whether the material earlier in the stretch was uncited with the result that I just didn't add the new material because I didn't think it responsible to add the citation to the material in front of to my new material without knowing for sure that the citation supported it. Sometimes you can edit around or work around that issue and sometimes you can't. This problem really needs a software solution so that when you add a citation you precisely indicate the material that it supports, which would support the goals of everyone on both sides of this issue. But with us now supporting two edit interfaces I'm not going to hold my breath for that one. FWIW, I teach Wikipedia editing classes 4-6 times a year and I strongly suggest to my students that they cite at least every sentence to avoid having their material deleted for being uncited if someone inserts new material in its midst. Regards, TransporterMan (TALK) 19:10, 9 May 2017 (UTC)

• Oppose. We don't need a citation after every sentence. The idea that a little superscript number after a sentence automatically makes it valid, which is what this breaks down to, is false. Firstly, there's the fact of the English language that ideas often take more than one sentence to fully express, and citations are there to provide verification of previously published ideas and facts. Secondly, regardless of how or where they're placed, there's always the possibility that a reference has been misinterpreted, or itself is made up. Verification rests on reading the references themselves, and just adding more little numbers pointing to the citation of a reference does nothing to aid verification. oknazevad (talk) 22:53, 9 May 2017 (UTC) (PS, the other aspect of cite overkill, that one sentence doesn't need too numerous citations, is also quite valid. Indeed, it's often a sign of SYNTH.)

This does not say we "NEED" a reference after every sentence. This is just saying that one of the three mentioned practices is reasonable (ie editors can reference every sentence if they want to). If an editor does not wish to reference every sentence than they should be prepared to do the maintain as "citation needed" tags or added or the content is deleted by new editors. Doc James (talk · contribs · email) 01:56, 10 May 2017 (UTC)

As I said above, editors should not be adding "citation needed" tags to, nor removing content from, properly cited material that is supported by the references. I am opposed to adding text to the guideline/essay that apparently gives editors license to do that. If editors are not bothering to check references, they should not be adding these tags in the first place, and should be gently notified of that. There should not be a guideline that apparently legitimizes inappropriate tagging and/or removal. Sławomir Biały (talk) 10:29, 10 May 2017 (UTC)

Have clarified to "may mistakenly" as the intent was not to suggest editors should just that they do. Doc James (talk · contribs · email) 00:31, 11 May 2017 (UTC)

Not everyone has access to all references. There is also no way of identifying whether the person who posted a "citation needed" tag has read and understood the reference that may or may not verify a sentence. When there are three or four references at the end of a paragraph containing a larger number of sentences, and someone tags one of them, it is quite a hassle to go through all of the sources to ensure that the tagged item can not be verified by any of the listed references, but that is what may have to be done to show that a tag is not appropriate. Particularly if they are large paper books and there are no page numbers and/or indices available. Gently notifying the tagger is probably even less productive than asking them if they have consulted all the references, and on which pages they did not find the disputed material. • • • Peter (Southwood) ^(talk): 16:34, 10 May 2017 (UTC)

User:Oknazevad, you said "We don't need a citation after every sentence." That is not what were are discussing here. When there is a citation available should it remain hidden or should the reader be able to click on the citation to verify the claim? QuackGuru (talk) 17:01, 11 May 2017 (UTC)

User:Pbsouthwood, you wrote "When there are three or four references at the end of a paragraph containing a larger number of sentences, and someone tags one of them, it is quite a hassle to go through all of the sources to ensure that the tagged item can not be verified by any of the listed references." No editor should be adding multiple sources at the end of a paragraph unless all the sources verify all the claims. Only the source that verifies each sentence should be placed at the end of a sentence. If a source verifies only part of a sentence then the source should be placed where it verifies that part of the sentence and should not be placed at the end of the sentence where it does not verify the entire sentence. See WP:CITEFOOT. Most editors do not understand where to place a source when it only verifies part of a claim. QuackGuru (talk) 16:59, 11 May 2017 (UTC)

QuackGuru, Are you saying that the current requirement is to have a citation to support the content of every sentence at the end of that sentence? Even when the reference is the same as for the previous sentence or sentences? I thought that getting agreement to allow this was the purpose of this discussion, (which I have not opposed). Also, unless I am mistaken, the citation is required to be after punctuation. As I understand it, the common practice of using a reference at the end of a paragraph is generally accepted as supporting any number of preceding contiguous sentences in the paragraph. Could be one, could be all, or anything between. Sometimes a whole section made up from several paragraphs, each with several sentences, is supported adequately by a single citation, other times a single compound sentence needs two or three, occasionally more. Then someone else may come and add another sentence into the existing collection, referenced by another source. It may or may not be supported by one of the existing sources. Does everyone check everything? Not a chance, even when it is possible. This is the reality of Wikipedia. Another day someone will clean up the broken flow and copyedit with some nice flowing prose which says the same thing in a different order and the article is improved, by one measure, but the referencing may be completely broken. It is a problem for which I have no easy solution, but referencing more sentences does potentially reduce the problem some of the time.• • • Peter (Southwood) ^(talk): 20:11, 11 May 2017 (UTC)

There are readers who think the text is unsourced if there is only one source at the end of a paragraph. That is a fact. The way to fix it is to add the visible source at the end of each sentence. Commenting the source out is not fixing anything. I support adding a source to support the content of every sentence at the end of that sentence rather than at only the end of the paragraph. QuackGuru (talk) 20:29, 11 May 2017 (UTC)

• Support, 1. having all references shown. I support adding a reliable source to every challenged sentence on Wikipedia. Readers might think the text is original research per this. On controversial topics editors intentionally replaced sourced text with text that fails verification, especially in the lede. Having a source after each sentence may help to curb the proliferation of original research. The practice of hiding the ref citation does not benefit the reader. QuackGuru (talk) 01:44, 11 May 2017 (UTC)

• Oppose -- I liked EEng#s idea here. I agree with Montanabw. That section has been stable for more than five years and serves us well. If new editors are dumping in {{cn}} tags, then we can simply revert them and explain on the talk page that the material is properly cited and point them to it. I do not buy this argument that it is too much work to verify the material. Our rule is the material in our articles is verifiable and if someone challenges it, we need to be up to the challenge to defend the material. If no one can be bothered to verify that there is a source that support the material, then we'll just have to accept that the material can be deleted by someone who challenges it. I'm okay with that. We are dealing with this exact issue at Latur Municipal Transport. No one can find WP:RS to support it though numerous editors insist it is important and should stay. Per policy, if those editors can't find WP:RS and there is no verifiability then it should go--simple as that... even if there really is such a thing... --David Tornheim (talk) 09:27, 11 May 2017 (UTC) (revised 11:42, 14 May 2017 (UTC))

You say oppose but your argument means support. See this edit by a reader. You don't tell readers on the talk page it is sourced. Readers think the text is unsourced even when there is a hidden citation. Challenged text must be verifiable. That requires a visible source at the end of the sentence. There are readers who may disagree with your argument and worse don't understand the source at the end of the paragraph might verify the claim. It is harder to verify the claim if a source is only at the end of the paragraph because the reader would not know if that source at the end of a paragraph was intended to verify all the claims. I have noticed editors add multiple citations after a sentence because none of the citations verify the entire claim. Regardless of the outcome here, admins continue to refuse to enforce the rules. Maybe we can create a new policy that requires admins to enforce the rules or they will be blocked. QuackGuru (talk) 16:47, 11 May 2017 (UTC)

No. I saw that diff, and that argument doesn't convince me. The editor who deleted it ignored that the statement was sourced, so that editor can simply be reverted. When the editor went to delete it, the citation was no longer hidden. And if you read the rest of the discussion here, even without the hidden comment, the material in the sentence was still cited and verifiable, in that the citation came somewhere else nearby. It is harder to verify the claim if a source is only at the end of the paragraph because the reader would not know if that source at the end of a paragraph was intended to verify all the claims. It is harder, and I have at times fretted about, but it is not a big deal and not worth damaging readability if large chucks of material are all from the same source. If the reader or editor distrusts and claim and wants to try to verify the claim, they must read the sources cited in the paragraph (and even better look for others). If they can't be bothered to read the sources, they have no right messing with the article: that is pure laziness IMHO and will lead to poor quality articles. The only exception is if they have no access to the source (such as a book or material behind a paywall), and then it wouldn't matter where the citation is anyway, since they have no way to look at it to verify it. I have noticed editors add multiple citations after a sentence because none of the citations verify the entire claim. That's a different issue which was already addressed in the cite overkill essay before this discussion started in these sections: Wikipedia:Citation_overkill#How_to_trim_excessive_citations and Wikipedia:Citing_sources#Bundling_citations. When you say admins continue to refuse to enforce the rules.: which rules are you claiming are not enforced? --David Tornheim (talk) 18:39, 11 May 2017 (UTC)

You wrote "The editor who deleted it ignored that the statement was sourced, so that editor can simply be reverted." That editor was a reader. Rather than revert a reader we can update this page to avoid causing confusion. No logical argument has been made to continue to cause confusion to our readers. The IP thought the text was unsourced. When readers think it is unsourced even when sourced then we can update the rules for the citations to be visible. It s not helpful to bundle citations into a single footnote. That should be deleted from WP:CITEBUNDLE. The policies including verifiable are not enforced by admins. Admins even replace sourced text with original research. QuackGuru (talk) 19:00, 11 May 2017 (UTC)

I think you are making too much hay out of that single diff. That situation was easily dealt with by a revert. When that reader used visual edit, the reference was there, and the reader overlooked it and deleted it anyway, saying "contradictory". Contradictory to what? We don't know. When the reader deleted it, s/he could have tried to figure out what the comment with the source was about, but was too lazy: the answer was looking him/her right in the eye. If I saw 20 examples of this, I might believe there is a problem. This is the first time I have seen it. I wasn't even aware of the commented citations was a thing until I saw this thread. 90% of the {{cn}} or WP:OR tags I see are warranted and are of far greater concern to me than stuff that has citations already.

I would support making a way for these commented citations to become visible when someone hovers over the sentence. That might solve the problem. I can't say I am a big fan of comments in Wiki-text in general, because they are not visible to readers or editors, unless they open the wikitext. I'm not a big fan of "secret" information like this embedded in the article, unless it is more a comment directed at editors/coders about the choice or functioning of the code itself, especially when a choice made was unusual and might require some explanation that readers won't care about. Then it is not hidden. Hence, I don't think I have even put one comment in any article though I have been aware of the ability to comment in HTML for probably 20 years.

We'll have to agree to disagree about WP:CITEBUNDLE

The other issue was addressed below. --David Tornheim (talk) 03:19, 12 May 2017 (UTC)

• Wrong concept. What we should have is one or more citations supporing each assertion of an idea, fact, etc. In some cases (say, the estimated value of some constant) there may be multiple views, requiring multiple citations a single statement. Or a single idea might require several sentences of explanation, with only a single citation. (Though such explanations might be something to put into the note.) The need for citations does not always correspond to individual sentences. (More on this if I ge some time.) ~ J. Johnson (JJ) (talk) 18:05, 11 May 2017 (UTC)

• User:J. Johnson, you say "In some cases (say, the estimated value of some constant) there may be multiple views, requiring multiple citations a single statement." Adding different citations that come to different conclusions at the end of a single sentence would make most of the citations fail verification unless all the citations verify the same claim. If a single idea might require several sentences of explanation, then a single citation can be added to each sentence. A single citation at the end of a paragraph does not verify the entire paragraph. QuackGuru (talk) 18:34, 11 May 2017 (UTC)
• J. Johnson is spot on. This is another important reason not to make the change. --David Tornheim (talk) 18:43, 11 May 2017 (UTC)

• The page says "One can hide citations with <!-- --> to prevent confusion in the future..." No. That is the causing confusion. That is a fact. The edit summary was "Remove contradictory uncited sentence". You think the reader is wrong and you are right? When readers think there is a problem it can easily be fixed by reverting over and over again across Wikipedia articles? No. The way to fix it is to make the citations visible to the readers. QuackGuru (talk) 19:16, 11 May 2017 (UTC)

• Agree this is a different conversation. One should generally only have one idea or concept per sentence. Doc James (talk · contribs · email) 22:00, 11 May 2017 (UTC)

• When two sentences are combined together editors usually place both sources at the end of the sentence rather than place each source to verify each specific idea or concept. When both sources do not verify the entire claim they should never be placed at the end of the sentence. QuackGuru (talk) 22:16, 11 May 2017 (UTC)

QG: There is a question of style on whether citations should be done through out a sentence as needed ("verify[ing] each specific idea or concept"), or collected together at the end of the sentence. Whether it is "usually" done one way or the other I don't know (some journals go one way while others go the otherway), but that's a different issue. Let's leave that alone, and focus on those cases where a bunch of citations is deemed "overkill".

If I might re-phrase your statement: you seem to be saying that multiple (more than "both"?) sources should not used if they "do not verify the entire claim", and perhaps we could extend this statement with "or contradict each other". However, those are exactly the cases where multiple citations may be needed. E.g., certain physical constants, the age of a possible artefact, the length of a certain earthquake fault, are likely to have different estimates of their value. To simply pick one that we (as an editor) like is a misstatement of human knowledge and a disservice to the reader. What we usually state is that there is a range of values, citing either a secondary source that "there is a range of values", or (good secondary sources often lacking) cite several experts showing the range of values. It is not a matter of two (or more) experts conflicting, therefore we can't use either; it's a matter of a conflict, and (several unstated caveats here) we describe the conflict. Stating that "values run from x to y, with a cluster around z" could easily require a half-dozen or more citations.

I was taught the same way as Doc: each sentence deals with one idea or concept, which get woven together to form more complex ideas. If a sentence is too complex it should get teased apart; if it has too little information content then maybe it is incomplete, and needs to be augmented, or joined to another sentence. But (as Doc said): generally. An explanation of an idea or concept might take several sentences, all from the same page or two of a text, and one citation suffices. Alternately (as I just showed), sometimes a simple "fact" might need a whole pile of citations. ~ J. Johnson (JJ) (talk) 00:19, 12 May 2017 (UTC)

If different sources come to different conclusions no editor should bundle sources at the end of each sentence. Where sources contradict each other multiple citations can be used. But the citations should be placed only where they verify each specific claim.

See "E-liquids are also sold without propylene glycol,[1] nicotine,[2] or flavors.[3]" Each citation verifies a specific statement. If all the citations were placed at the end of the sentence then it would be advisable to fix the problem or add a failed verification tag to each citation that does not verify the entire sentence. Multiple citations can be used in a sentence but each citation should be placed where they verify each statement. If sentences are joined together each citation must be placed where they verify each statement.

Bundling citations at the end of each sentence means each source verifies the entire claim. If they do not verify each claim then it can be fixed per WP:CITEFOOT. Commenting out citations as well as misplacing citations is an ongoing problem. One citation at the end of a paragraph is not enough. It improves the verifiability of the content when each sentence has a citation at the end. QuackGuru (talk) 00:54, 12 May 2017 (UTC)

I see the problem: You are trying to impose a standard that we do not have: But the citations should be placed only where they verify each specific claim. Although I do try to do this and see how it is helpful for anyone (including me) trying to verifying specific claims, I have never seen such a hard and fast rule. As J. Johnson has correctly stated this is not a requirement that is agreed on for publication. In other well-respected encyclopedias like Britannica, sources are often listed at the end of the article with no footnotes at all. For example Enlightenment. Now I see why you think admins are not following citation rules: They are not following a rule which you would like to exist, but which does not exist. --David Tornheim (talk) 01:38, 12 May 2017 (UTC)

I never said admins are not following citation rules. You are putting words in my mouth. They are not following basic rules including verifiability. Admins support editors who replace sourced text with original research and admins also replace sourced text with original research. You can quote me on that.

You wrote "Although I do try to do this and see how it is helpful for anyone (including me) trying to verifying specific claims, I have never seen such a hard and fast rule." It is not about trying to verify specific claims. It is about not placing citations where they do not verify the claims. Verifiability is about being able to verify each statement. That's not going to help verify each statement when citations are commented out or misplaced. QuackGuru (talk) 01:53, 12 May 2017 (UTC)

In all fairness, I have seen that behavior too, and it bothered me as well. Before I posted the below I was about to comment that in my legal writing course(s), the instructor(s) did say, "Legal writing is very different than other writing. Unlike in your other non-legal courses, you need a citation at the end of every sentence when stating the law." So I think I picked up the habit of using the legal type of citations here on Wikipedia, even though, I have always seen the advantage in readability of not having a citation at every single sentence, which is why I oppose the change. The ABA attorney discusses this issue at some length in the essay below. --David Tornheim (talk) 02:34, 12 May 2017 (UTC)

After a citation is placed after every single sentence for a paragraph an editor comes along and removes the citations. They think one citation is plenty. I don't understand how it impairs the readability having a citation after every sentence. I have some reading to do below and will be editing another essay where I will not get reverted. QuackGuru (talk) 03:25, 12 May 2017 (UTC)

• Comment I did a google search on "citing every sentence" and found some interesting articles discussing this issue.

• Legal Writing: ABA article about how too many citations clutter writing
• APA: [Library explains APA rules]: "APA demands that your reader knows exactly what information you got from someone else and when you start using it.Thus, an end-of-paragraph citation does not tq|meet that requirement." [emphasis removed]
• Chicago Manual of Style: FAQ from their website: "Q. When doing footnotes, do you put a footnote after every sentence... A. Footnotes should be placed where you need them, not according to a rule... If everything in a paragraph is from the same source, however, it’s enough to put one note at the end of the paragraph."
• MLA: [2] "Because citations are designed to show readers which material came from a source and which from you, it isn’t sufficient to cite once at the end of a paragraph."
• History Department Univ. British Columbia: [3] "Do I have to include a footnote after every sentence? No, if you include several facts or ideas within a single paragraph, they may be more conveniently cited at the end of the paragraph in a single footnote or endnote."

There may be more agreement than is apparent from the quotes: There seems clear agreement about when it is not acceptable to use an end of paragraph citation, specifically when the reader might wrongly confuse the author's ideas with those of the cited work.

--David Tornheim (talk) 02:25, 12 May 2017 (UTC)

• Support in a wiki, what is supported by a ref is only the sentence before the ref. Of course it's ugly when citations are everywhere (but not overcite imo, overcite is when there are too much refs for 1 sentence) and to avoid that I try to make long sentences supported only once but they're often cumbersome. But meaning (semantics) trumps form, multiple refs are better than unsupported text. The best way would be to have the supported text explicitly related to a ref as proposed before, like with a <referenced>...</referenced> tag, for example highlighted when you mouseover the ref.--Marc Lacoste (talk) 07:58, 12 May 2017 (UTC)

This is not consistent with Wikipedia:Citing sources, where it says that a citation may support a "clause, sentence, or paragraph", and discusses in multiple places adding citations to the end of a paragraph. Sławomir Biały (talk) 09:42, 12 May 2017 (UTC)

• Oppose. While I share some of Doc's sentiments, I find this proposal not well worked out. There is also entanglement of several sub-issues which makes the discussion more challenging than I have time or patience for. ~ J. Johnson (JJ) (talk) 22:33, 12 May 2017 (UTC)

• Comment I think the content should be replaced with something with more nuance. overkill is subjective. The reason mentioned for using less footnotes is readability. I think that in general adding a footnote after every sentence improves readability. It does not use much space while readers can directly see what source the information is coming from. If there only is one footnote at the end of a paragraph the reader cannot know if only the last sentence has a source, or the whole paragraph is sourced. Adding ambiguity does not improve readability in my opinion. --VENIVIDIVICIPEDIA^talk 15:35, 13 May 2017 (UTC)

Yes. Wikipedia is not like various journals where there is a high degree of competence and implicit trust between authors, editors, and readers. We have a greater need to show our readers – who are not experts, and often ill-informed, but perhaps properly dubious in that our editors are also generally not experts – exactly "what source the information is coming from." It is not enough that we show what (in our opinion) is, we absolutely must show why that is likely correct. Sources, sources, sources! ~ J. Johnson (JJ) (talk) 20:31, 13 May 2017 (UTC)

Nightscream, I've often seen you do this (cite WP:REPCITE). So do you have an opinion on any of this? Flyer22 Reborn (talk) 16:41, 18 May 2017 (UTC)

Strong Oppose. I support the current text of WP:REPCITE, as there is no justification for the inane citation clutter caused by putting a cite after each sentence. None of the arguments I've read above justify such a thing, and others have already stated why, so I will just refer to them, rather than engage in repetition [which is ironically consistent with my position on citations, if you think about it! :-) ]. Thanks for inviting me to this discussion. Nightscream (talk) 17:30, 18 May 2017 (UTC)

Nightscream, no problem. I meant to ping you earlier since I see you cite this essay a lot. QuackGuru went ahead and added a "citation needed" tag after seeing my above ping, though. I haven't yet checked to see if the whole sentence is already supported by one of the sources. If it is, then QuackGuru either didn't check the sources or he was making a point with that "citation needed" tag. Flyer22 Reborn (talk) 17:48, 18 May 2017 (UTC)

• Support - medical content really does need citations for every sentence. WP:V and the reliability of Wikipedia are more significant in this context than in pretty much any other type of article IMO; citing every sentence facilitates WP:V and enhances the reliability of medical content by clearly indicating whether a sentence is supported by a particular reference or not. Consequently, when I read medical articles, I assume that sentences that are not immediately followed by a citation are not cited. I've also found it to be really annoying when this essay is the basis of an objection at FAC in relation to the medical articles I've nominated in the past (e.g., amphetamine, hydroxymethylbutyrate). Seppi333 (Insert 2¢) 03:15, 26 May 2017 (UTC)

Citations: extended comments re "spanning"

• Comment I haven't the patience to read the above -- maybe someone said this already, but... somewhere, I'm sure of it, there's some facility that allows the "span" of text covered by each cite to be defined in the wikitext. I even think the reader can hover over the superscript cite callout to make the portion of text the cite covers light up in color. This must be buried in one of our 23 conflicting and overlapping help pages on citation mechanics. EEng 16:52, 13 May 2017 (UTC)

Perhaps you are thinking of {{abbr}}. I don't believe we have any "spanning" form of citation. ~ J. Johnson (JJ) (talk) 20:36, 13 May 2017 (UTC)

No, {abbr} has nothing to do with it. I figured it out. It's {{Citation needed span}}. Turns out it's for citation needed, not citations, but you can see what I was thinking of. If there was some way to tie a cite to a definite block of text that the cite covers, so that the reader could ask to see the boundaries of that block on demand, that might be the right idea. The problem, of course, is that there's no way to prevent someone from sticking something new in the middle of such a block -- something which is not, in fact, covered by the cite.

In very complex articles where this kind of issue was a problem, I've sometimes used the technique of adding a newline (not visible in the rendered page, of course) after each citation. That helps editors see at a glance the text block which immediately precedes, and therefore is presumably covered by, by that citation. EEng 21:17, 13 May 2017 (UTC)

@EEng:If there was some way to tie a cite to a definite block of text that the cite covers -- YES. Exactly what I was thinking of too. I would support opening up a new section or possibly something at Teahouse technical about seeing how hard it would be to get a draft template to do this--unless you think it would be easy. Who would have to approve it? Who wrote the code that does the hover stuff we do have? I just noticed it started happening one day... I've sometimes used the technique of adding a newline (not visible in the rendered page, of course) after each citation. That sounds good too--but I would think the newline would mess up the rendered page. I like the idea. Can you give us an example where you used it? --David Tornheim (talk) 11:37, 14 May 2017 (UTC)

Designing a template will involve many challenges. But the newlines-to-make-blocks-clearer is easy. Just try it somewhere. Except in some special places like bullet lists, adding a single linebreak in the middle of the wikicode has no effect at all on the rendered page. (Two linebreaks in a row is interpreted as a paragraph break.)

Adding "internal" linebreaks like this has an additional benefit: it makes diffs of complicated paragraphs more focused and easier to interpret. EEng 12:04, 14 May 2017 (UTC)

Thanks. I know what you mean about adding line breaks to help out the diff program. I was just doing that tonight, trying to make a minor change from looking like big change. You sound like a coder yourself (like me), esp. with the title EEng. Do you know if there is a spec. somewhere, like a grammar and syntax guide that says exactly how the machine interprets the wikitext? I don't have the energy to read the code directly (what language is it in?), which I imagine is huge (maybe not?), but a specification would be nice to answer questions like these. I find it very odd that carriage return follow by a space, gives a big multi-line block in light color. For years I would see that and it would take a while to figure out what went wrong. I find that completely bizarre choice, and have never seen it explained anywhere. And the same with what you are saying about line feeds. Even the way it interprets the opening [, [[, {{, [[:, [[File:, etc. seems to put it in a special kinds of parsing mode, each with unique rules, that often do something I do not expect, and a nice grammar definition would be easier than trying to read every single help page. It's got to be far more confusing to people who have never written a parser and don't understand computer language formal grammars. Maybe I should ask as Tea House Technical if you don't know? --David Tornheim (talk) 12:46, 14 May 2017 (UTC)

I doubt there's a formal specification anywhere, just a pile of code you can look at, and I predict you will find it a shocking and frustrating experience. Someone at WP:Village pump (technical) can no doubt point you in the right direction. EEng 13:18, 14 May 2017 (UTC)

This is the right thing to do.--Marc Lacoste (talk) 06:57, 15 May 2017 (UTC)

I've heard that there's no formal specification for wikitext (and that most of MediaWiki is written in PHP, so perhaps this is, too), but you all might be interested in what the mw:Parsing team is doing. The "old parser" is being replaced, and there might even, someday, far in the future, become something sort of like a specification. WhatamIdoing (talk) 01:51, 27 May 2017 (UTC)

Before anyone gets too deep into designing something new: a "spanning" type of citation has really quite limited use. (E.g., wrapping some text in a {spanning-citation} template precludes putting the citation in a note.) More useful would be a "spanning" type of note (specifically, that produced with <ref>...</ref> tags), which could then contain all of the citations and comments pertaining to the "spanned" text.

But the problem hardly needs to arise. E.g., where sentence contains one or more simple assertions (like, say, some number) it is simple enough to put the pertinent supporting source directly following that material. (The problem here is where some editors insist on moving all notes to the end of the sentence.) For more complex assertions, or where a source applies to multiple simple assertions, it is straight-forward enough to say as much in a note. ~ J. Johnson (JJ) (talk) 20:49, 17 May 2017 (UTC)

• Appeal to common sense: Cite sources logically. Citing is necessary, an I don't think the proposer or anyone else is arguing against citation, but not citing the same source repeatedly in sequential sequences. Something like This is a sentence.[1] This is another sentence.[1] This is a third sentence.[1] just looks unprofessional and rather silly. Where an entire paragraph is cited to the same source, and then some tag-bomber shows up, there is no reason not to do something logical like add hidden texts stating that the situation is, i.e.: This is a sentence. This is another sentence. This is a third sentence.[1]<!--entire preceding paragraph sourced to fn 1-->It is absurd to have a mechanical "cite every sentence" rule where the source is identical. That said, there is the opposite situation too: sometimes multiple sources may be needed in a single sentence : This is a clause,[1] with another, more detailed clause,[2] concluding with a third clause.[3] Yes, {[u|EEng#s}} is right that sometimes there is creep over time and content can get separated from sources, but that is what watchlists are for. Montanabw^(talk) 07:59, 14 May 2017 (UTC)

Agree. --David Tornheim (talk) 11:37, 14 May 2017 (UTC)

I disagree per this. The purpose of policy and guidelines is to try to slow does the majority of editors who are against the rules. Most people on Wikipedia should not be deciding what the rules say because they are against the rules. Editors continue to misplace citations. When multiple sources are needed for a single sentence each citations can be placed where it verifies the claim. If all the citations are placed at the end of the sentence then all the citations must verify the entire claim. Misplaced citations is continuing to cause confusion. If only one citation is placed at the end of a paragraph then any editor can delete any sentence that is unsourced. We need counter measures to improve the verifiability of content. It won't improve the readability of content when the text could be deleted. QuackGuru (talk) 15:23, 14 May 2017 (UTC)

What?? That is a distinctly odd statement, that "[t]he purpose of policy and guidelines is to try to slow does the majority of editors who are against the rules. ..." It might be interesting, perhaps even useful, to get into that, but also off-topic. And it distracts from the rest of your comment.

As to the rest: yes, I agree that sources (notes) at the end of a sentence should be applicable to the whole sentence (unless otherwise qualified), and sources (notes) applicable to only some sub-section cited immediately following that sub-section. Note that the same principle can applied to paragraphs.

Re Montanabw's objection: Would you repeat the same material in successive sentences? (E.g.: "Fact A first time.[1] Fact A second time.[1] ...") I think not; you would have different facts ("A", "B", ...). Even if all those facts come from the same source, they are likely to have different in-source locations (e.g., page numbers). So while you are correct in that all these facts are supported by the same source, they are not the same citation. The proper citations would be something like "Smith, 2001, p. 17", "Smith 2001, p. 24", "Smith, 2001, pp. 31-33". Something applicable to all of these can be added to the last note. E.g.: "While Smith 2001 has long been considered authoritative (Jones, 2012), a recent study (Brown, 2015) suggests the true values are 10% higher."

In summary:

1) It's not "overkill" to cite everything that needs to be cited.

2) Repetition of identical citations is a possible indicator of incomplete citation.

3) Issues of "spanning" can be handled at the meta level; syntactical level tools are not needed.

~ J. Johnson (JJ) (talk) 20:59, 17 May 2017 (UTC)

Flow regarding the content about hiding citations

QuackGuru, regarding this and this, I argue that the content flows better by stating what hiding a citation entails and then noting that it can be problematic. It does not flow well to begin by noting that citations can be hidden, then stating that it can be problematic, and then explaining the matter of hiding citations and then explaining how hiding them can be problematic. It's why I made the edits I did, including this and this one. Furthermore, stating that one can hide citations and then adding "but realize" is poor flow because even though you are trying to state that one should realize that hiding them can cause a problem, it can be read as "One can hide citations but then realize the problem." Flyer22 Reborn (talk) 20:24, 26 May 2017 (UTC)

"One can hide citations with <!-- -->." Our readers disagree. Editors may think the content is unsourced. It is time to move forward and fix the wording. QuackGuru (talk) 20:26, 26 May 2017 (UTC)

Well, I like I stated in the edit history, it is Doc James who added that bit. So I'm thinking he should weigh in on this. Flyer22 Reborn (talk) 20:29, 26 May 2017 (UTC)

One can hide all the citations throughout the entire article? It needs to be fixed. I never seen a page where all the citations are hidden. QuackGuru (talk) 20:31, 26 May 2017 (UTC)

The suggestion is simply that if one paragraph with three sentences is supported by one source, the first two instances of the ref for the first two sentences can be hidden with only the third and final ref of the paragraph showing. Doc James (talk · contribs · email) 00:27, 27 May 2017 (UTC)

That's not what it says and even if it said that it is a problem hiding citations per WP:INTEGRITY. QuackGuru (talk) 00:36, 27 May 2017 (UTC)

I agree editors on this talk page clearly have different opinions about this whole thing. That means there is no consensus for hiding citations. I don't see any article benefiting from hiding citations. QuackGuru (talk) 15:14, 27 May 2017 (UTC)

That does not mean that this needed to be added. You act like the text was stating that one must hide citations. All it was/is stating is that citations can be hidden. There is no need to state that there is no consensus on the matter. And regarding your WP:INTEGRITY argument above and this, WP:INTEGRITY is not about hiding citations at all. It is about making sure that the source used to support a sentence or paragraph actually supports it and has not been moved to an area it does not support. Flyer22 Reborn (talk) 15:20, 27 May 2017 (UTC)

It is against hiding citations. See WP:INTEGRITY: "The point of an inline citation is to allow readers and other editors to check that the material is sourced; that point is lost if the citation is not clearly placed." The citation will be misplaced if they are hidden and they will be lost for our readers. The point is to let the reader verify the claim. Please show me an article that benefits from hiding citations for our readers. QuackGuru (talk) 15:26, 27 May 2017 (UTC)

I've clearly seen WP:INTEGRITY; I simply do not agree with your interpretation of it as far as hiding citations go. It seems you need to read the rest of that guideline, including its examples of what WP:INTEGRITY violations are. I see no issue at all with hiding some citations in the case of a citation overkill matter. The citations are not misplaced by being hidden; they are simply hidden. Unless you can show that all of the citations need to be displayed, I will not be agreeing with you that hiding citations is inherently bad. Hiding them can be problematic, I agree; I do not agree that hiding them is automatically bad. And since you often interpret our policies and guidelines in ways that others do not, as also indicated above on this talk page, I do not think it is beneficial that I continue to argue this matter with you. Flyer22 Reborn (talk) 15:38, 27 May 2017 (UTC)

I'm going to bump it up to a policy page. See WP:VERIFY: "All content must be verifiable. The burden to demonstrate verifiability lies with the editor who adds or restores material, and is satisfied by providing a citation to a reliable source that directly supports the contribution.[2]" It says "All content must be verifiable." Our readers cannot verify the claim when the citation is not visible. It is clearly against policy to hide citations. If you disagree then explain how our readers are going to verify a claim when there is no visible citation. QuackGuru (talk) 15:46, 27 May 2017 (UTC)

The policy also says that verifiABLE "means that a source must exist for it, whether or not it is cited in the article." That gap between "able to be verified, through any means, including by doing your own web search or going to your own library and asking your own reference librarian for help" and "someone has already typed a citation to a source (that is allegedly reliable and allegedly supports the material in question) into the article" seems to trip you up fairly often. The policy requires that some people be ABLE to verify the material. It does not, and never has, required that everything be CITED already. WhatamIdoing (talk) 03:19, 28 May 2017 (UTC)

The policy says under Wikipedia:Verifiability#Original_research: "means that a source must exist for it, whether or not it is cited in the article."

What you quoted what not about verifying a claim. It is about a different, unrelated matter regarding original research.

For years it says "This means all material must be attributable..." Verifiable does not mean as long as a reader can verify the claim somewhere on the web or at a library the text is verifiable. Verifiability is accomplished by providing a citation according to WP:PROVEIT.

According to your argument as long as the text is verifiable a citation is not even needed. QuackGuru (talk) 03:31, 28 May 2017 (UTC)

Indeed, that is the policy: Material (text or otherwise) does not need to have a citation, unless the material falls into one of the four categories in WP:MINREF (three of which are determined by WP:V, and the fourth by BLP). The policy is not verifiED; the policy is verifiABILITY. If someone (i.e., anyone other than the editor who originally added the material) has the ability to verify that the material could have come from a reliable source, then the material is verifiABLE. The presence or absence of citations in the text is irrelevant. The policy only demands that it be possible, not that it already be documented. It is perfectly allowable under WP:V to have a sentence such as "Smallpox is a vaccine-preventable viral disease" without a citation after it. That material is verifiABLE to anyone who spends a few seconds with their favorite web search engine, and therefore it fully complies with WP:V even if it is uncited. WhatamIdoing (talk) 08:25, 28 May 2017 (UTC)

I challenge all unsourced statements. Any statement on Wikipedia that is unsourced is now challenged. QuackGuru (talk) 15:36, 28 May 2017 (UTC)

Including "the human hand normally has five digits", which even young children should know?

Seriously, editors have tried that kind of end-run around the policy before, and the community rejects it as WP:POINTY and disruptive. WhatamIdoing (talk) 17:01, 28 May 2017 (UTC)

But we do cite a source for the human hand having 5 digits. See Hand: "The human hand has five fingers and 27 bones, not including the sesamoid bone, the number of which varies between people,[3]..."

Many editors reject policy and follow their own rules. QuackGuru (talk) 18:34, 29 May 2017 (UTC)

That is often OK until someone objects, at which point policy tends to have the upper hand until changed. • • • Peter (Southwood) ^(talk): 13:57, 30 May 2017 (UTC)

Policy is not consistently enforced. If it was enforced then editors would stop replacing sourced content with text that fails verification. QuackGuru (talk) 14:02, 30 May 2017 (UTC)

Of course it isn't, Even the law isn't consistently enforced, and there are professionals trained and employed to do that A deviation from policy must be noticed by someone who knows that it is a deviation from policy, considers it an actual problem, and cares enough to do something about it, before anything is likely to be done. Then there is the problem that it is quite common for several people to disagree with any given policy, and with the interpretation of that policy regarding the current application, which can slow things down. Even if the specific policy was enforced, failing verification is also a minefield of uncertainty. It can happen simply because the source has changed, it can happen because what one person understands as what the citation says may differ from another person's interpretation, and as you are aware, we are required to restate the information to avoid plagiarism. It can also happen because the wrong source was cited, intentionally of by accident. It is a problem. • • • Peter (Southwood) ^(talk): 15:35, 30 May 2017 (UTC)

There are numerous articles that contain text that fails verification. For example, read the chiropractic talk page. For almost ten years I have been trying to remove the failed verification claims. Now it is only one paragraph in the lede. QuackGuru (talk) 15:50, 30 May 2017 (UTC)

I would not be surprised if the majority of articles contain text which would fail verification. One of the problems with failing verification is that it does not prove a statement to be wrong, hence the "verifiability not truth" principle. In a way this is less of a problem with uncited material, as people are more inclined to assume that a referenced statement is verifiable without bothering to check, so a bad citation can be worse than none at all. Shit happens, eventually much of it gets cleaned up. Sometimes it takes years. On Wikipedia we can watch it happening, elsewhere we may have to assume good faith because there may not be an option.

What is the root cause of these problems? The Wikimedia Foundation (WMF)? The WMF ignores almost everything on Wikipedia. QuackGuru (talk) 16:20, 30 May 2017 (UTC)

This is a wiki edited by volunteers. No-one is obliged to check references. The wiki is growing constantly. That some references remain unchecked and incorrect is a logical consequence of those conditions, and nothing to do with WMF, which has no obligations to check or interfere with the development of the encyclopedia providing that editors comply with the terms of use. • • • Peter (Southwood) ^(talk): 07:59, 31 May 2017 (UTC)

QuackGuru, although I don't fully agree with WhatamIdoing on verifiability, it is not true that everything needs to be sourced. What needs to be sourced is material that is likely to be challenged. Common sense material does not need to be sourced. We absolutely do not need to source that people have five fingers on each hand. And as for hiding citations, my point is that it's not always the case that all of the citations need to be shown in the article. What is important is that the sentence or paragraph is clearly supported by at least one reliable source. Flyer22 Reborn (talk) 21:24, 30 May 2017 (UTC)

When it can be sourced, then "attribute to" a source rather than claim it is "verifiable in" a source and not provide a source for the claim. Hiding citations will be hiding citations from the reader. The reader cannot verify the claim without a citation. It is more important to verify the claim for the reader. WP:INTEGRITY is another important point. After a sentence is deleted, editors continue to leave behind the citation where it does not verify any claim. QuackGuru (talk) 23:57, 30 May 2017 (UTC)

You stated, "Hiding citations will be hiding citations from the reader. The reader cannot verify the claim without a citation. It is more important to verify the claim for the reader." Did you not grasp what I just stated? I am talking about cases where the claim is already verified by one or more sources. In cases where content is already verified by one or more sources, the additional sources do not need to be shown; they can be hidden, just like they can be bundled. Why hide them instead of remove them? Maybe because they are or will be useful in the future. And, no, we don't have to cite everything; I've already noted why and will not be repeating myself to you on the matter. Flyer22 Reborn (talk) 02:14, 31 May 2017 (UTC)

When there is no visible citation at the end of a sentence the reader cannot be expected to know here to find a citation to verify the claim. Maybe the citation at the end of the section verifies the claim or maybe it does not. An editor may start tagging the sentences without a citation. I have no problem citing everything I add. It is not hard for me. Verifiability is far more important than allegations that the content will be less readable due to one citation after each sentence. QuackGuru (talk) 02:23, 31 May 2017 (UTC)

I agree that verifiability is important, and that it should be possible to see a useful citation, but also that some statements will never be challenged, and therefore do not need a citation. However there is no fundamental harm in providing a citation for something which may be, as Al Capp so pungently expresses, "as any fool can plainly see". Most of us are happy to wait until some exceptionally under-informed person challenges an obvious statement before providing a citation, but, as is often the case with medical articles, if the editor sees fit to provide a citation for every statement to reduce the hassle of looking it up again later, then good for them, and it should be accessible to the reader in some useful way. This should be permitted, but not obligatory. It would be nice if there was a software method of hiding references during reading. For example, a toggle button to hide/show references. Ideally this would remain visible at all times, which might be a technical challenge for mobile view. For desktop maybe a button in the sidebar that moves down the page with the reader, for mobile, maybe an icon at the top of the screen. • • • Peter (Southwood) ^(talk): 07:59, 31 May 2017 (UTC)

QuackGuru, it's clear that you do not understand what I am trying to state about hiding citations. I repeat: "In cases where content is already verified by one or more sources, the additional sources do not need to be shown; they can be hidden, just like they can be bundled." If a sentence in an article is already verified by a visible source and the additional sources are not needed to verify that sentence, they can be hidden. An editor might hide those sources in the article because they can be used for something else later in the article. So, no, those additional sources, which are causing citation clutter, do not need to be visible in the article. Flyer22 Reborn (talk) 14:46, 31 May 2017 (UTC)

This discussion is not about adding additional citations or hiding additional citations after a sentence when there is already one citation after a sentence. This is about when there is no citation after a sentence and there is only one citation at the end of a paragraph. If there is only one citation at the end of a paragraph the other sentences that don't have a citation need a citation. Those citations should not be hidden because the reader can't verify the claim when they are hidden. Bundling different citations into one citation is very confusing to editors and to readers. QuackGuru (talk) 15:10, 31 May 2017 (UTC)

This discussion concerns the "Needless repetition" section, which is about trimming needless repetition and includes the "hiding citations" aspect that you disagree with. As for citation bundling, I doubt that it is "very confusing to editors and to readers." I sometimes use WP:CITEBUNDLE and using it is allowed. You should take any cite bundle issue you have to that guideline talk page. Flyer22 Reborn (talk) 15:49, 31 May 2017 (UTC) Flyer22 Reborn (talk) 15:52, 31 May 2017 (UTC)

It is not "needless repetition" to make visible a citation for the reader and if the citation is not visible the reader will not be able to verify the claim. That makes the text unsourced to the reader. As for citation bundling, it is misleading. It makes it more difficult to verify a claim. Which source verifies a claim when different sources are bundled together? There are times when more than one citation is needed to verify the entire sentence. The best approach is to place each citation only where it verifies each particular claim rather than place all the citations at the end of a sentence where all the citations do not verify the entire sentence. There are other problems. I noticed there are editors who add multiple citations but all the citations do not verify any of the content. Editors are getting smarter to get around policy. Uninvolved admins most often refuse to enforce the most basic policy. A lot of things are allowed but that does not make it appropriate.

See Wikipedia:Verifiability: "In Wikipedia, verifiability means that other people using the encyclopedia can check that the information comes from a reliable source." Other people including our readers can't check to verify a claim when there is no citation at the end of each sentence. A citation way down at the end of a section without providing visible citations for the other content is deemed to have failed verification for the other content according to Wikipedia:Verifiability policy because other people must be able to be given the opportunity to be able check to verify the content. The other content is considered unsourced to who? Our readers. Hiding citations is a flawed concept. QuackGuru (talk) 17:38, 31 May 2017 (UTC)

Really? Other people, including our readers, can't use their favorite web search engine to determine whether facts have been published in a reliable source? Other people can't go to the library? Other people can't buy and read books? The phrase that says "can check that the information comes from a reliable source" includes every possible method. It does not require that other people be able to "check that the information comes from a reliable source" through the exclusive method of clicking on a blue number and looking at that citation alone.

Also, there should not be any statement anywhere in Wikipedia that says humans normally have five fingers. Strictly speaking, each human hand has five digits, or four fingers and one thumb.

The citation at the end of that slightly wrong sentence has nothing to do with the number of fingers on the human hand. The citation was added with the material about 27 bones. The bit about the number of fingers was added later[4] and cannot be assumed to be supported by the source. I've split them back into two sentences, and described it with more precision. WhatamIdoing (talk) 16:01, 3 June 2017 (UTC)

QuackGuru, I agree to disagree with you. Also, there is no need to point me to policies or guidelines I am aware of.

WhatamIdoing, depending on the source/definition, the thumb may be defined as a finger. This is even currently mentioned in the Finger article. And I state "currently" because I see that QuackGuru has a habit of going to articles and changing them when they are highlighted in a discussion; I see that he also followed you to the Hand article. Flyer22 Reborn (talk) 01:36, 4 June 2017 (UTC)

Disputed template

I have placed a disputed template on this essay as "repeating references" is prefered by many. Doc James (talk · contribs · email) 18:58, 8 June 2017 (UTC)

It is disputed. The wording can continue to be improved. QuackGuru (talk) 19:42, 8 June 2017 (UTC)

Citations are not "Needless repetition". We may need to change the name of this essay. QuackGuru (talk) 23:18, 8 June 2017 (UTC)

It is when it it's excessive and unnecessary. Flyer22 Reborn (talk) 23:29, 8 June 2017 (UTC)

I reverted QuackGuru on adding "this is disputed" throughout the essay after Doc James added Template:Disputed to the top of the essay. It makes no sense to change "Material that is repeated multiple times in an article does not require an inline citation." to "It is disputed for material that is repeated multiple times in an article does not require an inline citation for every mention.", or "This is correct" to "It is disputed that this correct." I also reverted Doc James on this bit. This is an essay that is specifically about not excessively citing. If someone wants to create a counter essay about excessive citing being perfectly fine, then they are free to do so, in the same way that we have WP:Sky is blue and WP:You do need to cite that the sky is blue. Also, this is not an article; so Template:Disputed does not belong on this page. In the meantime, I will be starting an RfC on this matter and alerting as many pages as I alerted for the initial discussion above about changing this essay.

And do stop edit warring. Flyer22 Reborn (talk) 23:29, 8 June 2017 (UTC)

Should this essay be changed to encourage more citations?

In the "Citations" discussion above on this talk page, a suggestion was made to include text that encourages more citations. For example, in this discussion, it was suggested that adding a citation after each sentence instead of having one citation at the end of a paragraph supporting the entire paragraph is beneficial, and that the text in the essay that discourages more citations should be replaced with text that encourages more citations. One view is that the essay is specifically about citation overkill (excessive citations) and that it offers leeway for cases where it is best to cite multiple sources, but that excessively citing should generally be discouraged. The other view is that more citations are beneficial; for example, adding a citation for each sentence of a paragraph assures the reader that all of the content is sourced, since readers can otherwise assume that some text is unsourced.

Opinions? Flyer22 Reborn (talk) 00:03, 9 June 2017 (UTC)