Wikipedia talk:Citing sources/Archive 53

This is an archive of past discussions on Wikipedia:Citing sources. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 50

Archive 51

Archive 52

Archive 53

Archive 54

Archive 55

Archive 56

Referencing WP:CITEVAR, an editor in this edit converted a newly added reference using sfn to a citation removing the sfn. There is currently no consistent citation style in the article, and sfns have been used in the article since 2012, so it seems to me that CiteVar is not applicable. An existing ref was not being converted to non inline form. The sfns were for new references (to de Waal and Damasio).

My inquiry is whether WP:CITEVAR means that editors cannot use Sfn templates in an article they have been existed for over a decade? I recognize that adoption of a uniform style is desirable and would require a consensus, but I am unclear on what the guidance is for editors simply making individual edits using a citation style already in use in an article. Perhaps some additional clarification on how editors think incremental upgrades of articles ought to proceed. There is a lot of reluctance to engage in multiweek discussions attempting to gain consensus on where to place periods and commas, and a lot of articles with highly nonuniform and low quality citations. Placing obstacles for bold editors willing to put in the time move articles towards improved citations in my view is not helpful. J JMesserly (talk) 04:01, 6 March 2023 (UTC)

To be clear, we're talking about 2 uses of SFN in an article that otherwise had more than 150 inline citations. MrOllie (talk) 04:04, 6 March 2023 (UTC)

A review of the article and its history will reveal that there is no consistent citation style in place nor has there been in its history. So is an editor justified in claiming CitVar justifies banning use of particular templates if there has never been a consistent style, on the grounds that the template being banned has rarely been used in the past ? J JMesserly (talk) 04:22, 6 March 2023 (UTC)

2 out of 150 is very clearly in favour of whatever style the other 148 are. While you can use sfn to source things if you don't know how else to source things, just don't revert people that cleanup after you and bring the article inline with consistency. Headbomb {t · c · p · b} 06:15, 6 March 2023 (UTC)

Yep. Consistency here means not introducing more {{sfn}}s, and ideally, fixing those that do exist. XOR'easter (talk) 15:20, 6 March 2023 (UTC)

@Headbomb: not necessarily. Consider articles that maintain a section where contributors have come to an agreement on which the high caliber, most notable sources are listed, such as the Ludwig Wittgenstein article. The style in that article is to use an sfn to such sources in order to avoid making a redundant copy of the reference inline. This is consistent with the guideline which specifically suggests sfn or rp. Yet to understand your point of view, the minute anyone makes such a reference to such a highly regarded source, you would have them remove the source from the recommended reading section as XOR'easter did in this edit of the emotion article- performed when the two editors decided to unilaterally ban sfns from the emotion article. Would you ban sfns from the Wittgenstein article as well? What are you saying- that the guidance to avoid redundant copies of citations is wrong and that rp harv and sfn should not be used in this case? "Inconsistency" is a canard and misses the structure the contributors of Wittenstein are using. If you ref a notable source in the bibliography, you use sfn. If you have a source not in the notable section, you inline ref it. The value is that the article is not establishing a false equivalency from the highest quality sources from those which may be far less notable. You have to look at the style of citation that the contributors in an article have developed. J JMesserly (talk) 05:08, 7 March 2023 (UTC)

Reread what I wrote because it is nothing like what you understood. Headbomb {t · c · p · b} 05:38, 7 March 2023 (UTC)

I understand by "2 out of 150 is very clearly in favour of whatever style the other 148 are" that you believe that such counts are relevant to the analysis of articles like Ludwig Wittgenstein. I explained why they are not. Context of how the inline <ref> or how the sfn is being used matter. You think it is proper to ignore context in such an article. This is a case a style can be consistent and require both be used in those confined contexts. It is a coherent style and those who violate it would run afoul of citevar. See what I mean? J JMesserly (talk) 06:26, 7 March 2023 (UTC)

reference info for Ludwig Wittgenstein

unnamed refs	121
named refs	36
self closed	26
R templates	1
Refn templates	1
cs1 refs	165
cs1 templates	189
cs1-like refs	1
cs1-like templates	3
harv refs	2
harv templates	7
sfn templates	178
refbegin templates	3
cleanup templates	13
webarchive templates	2
use xxx dates	dmy
cs1|2 dmy dates	37
cs1|2 mdy dates	2
cs1|2 ymd dates	2
cs1|2 dmy access dates	38
cs1|2 ymd access dates	7
cs1|2 dmy archive dates	24
cs1|2 ymd archive dates	2
cs1|2 last/first	136
cs1|2 author	6
List of cs1 templates cite book (2) Cite book (86) Cite encyclopedia (1) Cite journal (28) Cite magazine (5) Cite news (18) Cite press release (1) cite web (4) Cite web (44)
List of cs1-like templates Cite IEP (2) Cite SEP (1)
List of sfn templates sfn (178)
List of harv templates harvnb (4) harvp (3)
explanations

The Ludwig Wittgenstein article is a mess, with no coherent citation style, and it should not be used as an example of anything except a page needing cleanup. There's no way to tell from reading it whether the listed "Sources" are actually more highly regarded, or whether they simply came first and later citations were added in using a different style. (A quick glance at even the recent editing history confirms the latter happening.) It refers to the same source using different styles. The structure is an illusion. XOR'easter (talk) 19:57, 7 March 2023 (UTC)

The Ludwig Wittgenstein article is a mess Confirmed. See the {{ref info}} rendering →

—Trappist the monk (talk) 20:21, 7 March 2023 (UTC)

WP:CITEVAR: Editors should not attempt to change an article's established citation style merely on the grounds of personal preference, to make it match other articles, or without first seeking consensus for the change. This come with a big warning that this has been strictly forbidden since at least 2006. WP:DUPCITES: Combine precisely duplicated full citations, in keeping with the existing citation style (if any). Hawkeye7 (discuss) 05:28, 7 March 2023 (UTC)

Hawkeye, No one is disputing that principle. I shift attention from the first example I gave to the sfn which previously existed in the article for 12 years and was eliminated in this edit. I had previously presumed the article had no consistent style in the past due to the typical cruft you see even in GA class articles. But I see now that for the contributors to the emotion article, no one had a problem with the sfn to the Further reading authoritative text for 12 years. There were ample numbers of highly capable editors who saw it and saw it as proper. Consider the possibility that maybe they were right. I see now that the emotion article was simply doing the same style that the Wittgenstein article is doing. Mixing the two is the article's style but it is not just any willy nilly mixing. Sfns are only allowed to texts in the recommended reading section. By removing the sfn the editors advocating an sfn ban disregarded the style and degraded the recommended reading section to achieve this end. If you look at the edit you will see that the authoritative textbook information was removed from that recommended reading section and placed inline, with the rationale that any consistent citation style must be either all inline or all non inline, and never the twain shall meet. I think that was wrong. The proponents point to citevar and claim they are making the article more consistent. Which is the correct position?

The reality is that many articles like Wittenstein have a style that successfully and I think properly mix them. I think that by banning the SFN from the emotion article the editors were in fact violating a perfectly acceptable style of citation. Just for clarity, from this perspective, the edit I first asked about would indeed be against this article style since it did not reference one of the highly notable texts recognized in the recommended reading section. But the existing sfn would in fact be required by the style and on these grounds the 12 year old sfn would properly be restored.

Let's make this into a practical question. Say the regular contributors to Emotion article want to keep the information on the Fox textbook in the recommended reading section. Should the guidelines say in this case that the cite should be duplicated both in the recommended reading section and in an inline <ref>? Or should the editor use a cite ref= field and then employ that name inline? But now you are just manually doing what rp harv and sfn do with far less syntax. Isn't it simpler to say that for such situations it is not inconsistent to allow an rp, harv or sfn?

My view is to best not to get into the weeds on these details or be prescriptive and simply say in the guidelines that using both inline refs and non inline mechanism like rp sfn and harv is permissible if they are used in a consistent way.

Can folks go along with that? J JMesserly (talk) 06:42, 7 March 2023 (UTC)

There is no evidence of this hypothesized large group of editors that has been agreeing with your position for 12 years. We've had typos persist in the encyclopedia for years as well. If someone typed 'the' as 'teh' and it took years to notice that doesn't mean that we have consensus for a mixed spelling style. Also, MOS:FURTHER is clear that citations should not link to the Further reading section. If you have seen that around Wikipedia you are seeing errors that should be fixed. MrOllie (talk) 13:26, 7 March 2023 (UTC)

Emotion is a garbage fire of an article, full of problems that nobody has had the time and/or will to fix. In other words, it's a typical broad-scope article. So far, I've found a crackpot source, a misrepresented and misplaced source, a source in the first sentence with broken formatting and missing authors, and three footnotes floating randomly at the end of a section, not attached to a damn thing. XOR'easter (talk) 20:05, 7 March 2023 (UTC)

I agree Mr. Ollie’s account may very well be the case for the two rare uses of sfn in the emotion article.

Note that banning hybrid mixture styles as proposed does also apply to articles such as Ludwig Wittgenstein which has both hundreds of sfns and hundreds of inline references. This usage has an observable rationale and a simple orderly rule: Make sfns only to notable authorities listed in a recommended reading section. Hypothetically, if the an article such as Wittgenstein removed the departures from that rule, would it be acceptable, or does the very mixture of inline with non inline reference make it a "mess" as XOR claims? There is a rationale here- note that less notable sources are not awarded a false equivalency by being located adjacent to highly regarded sources organized rationally and set aside from the chaotic hoi poloi of references to dissertation theses and popular press references. It is claimed the Wittgenstein citation style is against guidelines because MOS:Further states that the “Further reading” section is not intended as a repository for full citations used in the article. But consider the possibility the regular contributors there understand and respect the principle described- they certainly aren’t using it as the main repository for full citations. So their style is conforming to the intent statement of MOS:further.

Let’s stand back from this for a moment. Should our guidelines be saying that no article may adopt a style which accepts the concept that the article would actually make citations to the respected sources in the further reading section? If not, then how do we propose they both keep the notable source in the recommended reading section and also make reference to it without the rp, harv or sfn templates? By Mr. Ollie and XOR'easter’s edits, I presume what is being proposed is that the guidance means editors must do as they did- demote the notable source from the Further Reading section, and bury it in an unsorted mass of references to far less notable references including dubious sources.

Does this guidance make wikipedia better for our readers?

The Wittgenstein style is acceptable, and recognizes the that Further reading is not intended to be used as a repository for all citations. Most of the cites are indeed in the References section. But they do allow links to be made to high quality sources using sfn. I am not saying I prefer the Wittgenstein style or that I think its sources sections are not in need of improvement. But should the Citations guidance be taken to mean that such a style with a mixed use of inline and non inline citations inherently inconsistent and therefore to be discouraged?

I think the guidance should make some statement that makes it clear that no such banning of such article styles is intended- only that such hybrid mixture styles be consistent. Can everyone go along with some sort of statement along those lines? J JMesserly (talk) 20:23, 7 March 2023 (UTC)

There is a rationale here- note that less notable sources are not awarded a false equivalency by being located adjacent to highly regarded sources organized rationally and set aside from the chaotic hoi poloi of references to dissertation theses and popular press references. Peter Hacker, Rudolf Carnap, Bertrand Russell and Thomas Kuhn are in the hoi polloi.

Yes, it is a mess. Whatever structure you are seeing isn't there. XOR'easter (talk) 20:33, 7 March 2023 (UTC)

Please don't dodge the question of the hypothetical. Let's say your hypothesis is wrong and the structure is there. In that case, and if the structure was cleaned up to observe it, should guidelines still discourage any such mixture? J JMesserly (talk) 20:51, 7 March 2023 (UTC)

It's not a hypothesis; it's a plain observation. And yes, encouraging that kind of disaster area would be a terrible idea. XOR'easter (talk) 21:03, 7 March 2023 (UTC)

To be clear, what I heard you say is that even if the article had the structure I described, and the style's rule constraining when sfn was permissible was uniformly applied, that guidelines should discourage if not bar any style permitting a mixture of inline and non inline citations. Is that a fair understanding of your position?J JMesserly (talk) 21:57, 7 March 2023 (UTC)

To be a consistent style, it needs to actually be a consistent style. You keep asking "but what if I want an inconsistent style – can I declare that to be my style, and call it consistent?" Is that a fair understanding of your position? If so, the answer is no. —David Eppstein (talk) 22:44, 7 March 2023 (UTC)

I see the levity of your rhetorical question, but let's be serious. No of course it is a most unfair representation of my position, and no the style is that which has evolved in the Wittgenstein article- not my style- I had no hand in it. The structure of the segregation of types of sources is elaborate- the motivation appears to be that it is more about a visual style of segregating types of sources for the reader than it has to do with uniform application of wikitext for every single citation, which I don't think readers care anything about, nor is relevant to guidelines. But David I honestly wish to comprehend your position. It appears to be that having a mixture of inline citations and non inline citations is by definition inconsistent regardless if they follow a particular rule and rationale as I described to segregate sources. If I have left anything out please correct me. J JMesserly (talk) 22:51, 7 March 2023 (UTC)

When you say what the motivation appears to be in the Wittgenstein article, you are projecting a motivation into it. There's no rhyme or reason, just one editor after another doing what they felt like in the moment. XOR'easter (talk) 05:36, 8 March 2023 (UTC)

Take a look at World War II article. It was reviewed April 25, 2013 and was awarded GA status and it did so with 70 harv references used in the way I described. See the history here is a version from May 2013.

It seems to me that what is confusing people in the guidance is that the requirement for consistency in the citation visual style of the article implies some sort of uniformity in how references are encoded in wikitext (eg- that all references must be either <ref>s or sfns). I have shown a high quality article can have mixed inline and non inline references, so I think we have to accept the authority of the peer reviewers. Since there is so much confusion about what is permissible, the guidelines need some clarification on this issue. J JMesserly (talk) 06:17, 8 March 2023 (UTC)

It would be a consistent style to have any one of

1. References that are used only in one way (books used only once, web pages, or journal articles) have the reference listed in full in the footnote, but book references where different footnotes refer to different pages or page ranges are listed with a short footnote and the full reference in a later section
2. Some footnotes with full references, and other footnotes that use different pages from the same source listed as short footnotes pointing back to the footnote with the full reference
3. All footnotes give full references, with maybe multiple page ranges in the same footnote if needed.
4. All footnotes are short and all full references are in a later section.

In all of these styles, one can also have a "further reading" section that gives material that readers might find helpful but is not intended to be used as a reference for the article content. The emotion article's style, prior to your edit-warring, appears to be more or less in the third of these consistent styles, although the individual reference formatting and the information about page numbers from books could use significant improvement.

It is NOT a consistent style to have almost all references listed in full in the footnotes, as in #3 above, but then suddenly to have a single reference be formatted more like #4 above, despite it being used with only one range of pages and despite there being no obvious distinction between it and the other full-in-the-footnotes references, as you have been trying to do. It is also not appropriate to have the later full reference be in a "further reading", as you have been trying to do; that section heading means something different. —David Eppstein (talk) 08:25, 8 March 2023 (UTC)

David, thank you for explaining. But there are some misunderstandings here.

1. there was no edit war. All parties went to discussion when it was clear there was a major disagreement on principles.
2. the sfn to the further reading section was not added suddenly. It was established by another user 12 years ago as part of the same style being used by the WWII article. (see above link).
3. your rules on what styles are and are not consistent would have led the peer reviewers to not award WWII article a GA status. BUt they did.

J JMesserly (talk) 19:37, 8 March 2023 (UTC)

Although many GA reviewers look at the consistency of formatting of citations, this is actually not part of the Good Article criteria. Being a Good Article does not require meeting Wikipedia best practices in all areas. However, as part of MOS:LAYOUT, MOS:FURTHER is part of the Good Article criteria (both 1b and 2a). So even though consistent formatting of references is not strictly required, avoiding the use of further reading sections for publications that are used as sources actually is required. —David Eppstein (talk) 22:47, 8 March 2023 (UTC)

• TLDR, but I agree with "Consistency here means not introducing more {{sfn}}s, and ideally, fixing those that do exist" above. The most important principle in WP:CITEVAR regarding changing styles is to have a discussion on the talk page, rather than making undiscussed changes and then wiki-lawyering about it. Very many editors disregard WP:CITEVAR entirely when adding material (very often using sfn), & the fact that a few sfn refs have been in the article for a long time means nothing. Johnbod (talk) 22:36, 8 March 2023 (UTC)

• J JMesserly, having read most of this (and skimmed a few parts), you seem to be arguing against the theme of consistency which is the raison d'etre of CITEVAR, and you give specific reasons for it. Without either agreeing, or disagreeing with your position (or anyone else's), here's my 40,000-foot take: in the grand scheme of things, we serve our readers, first, while attempting to maintain order and uphold our principles, of which in this case, the most notable is the core principle of WP:Verifiability. I think it's probably uncontroversial to say that both you, and editors that may disagree with you here, both believe that verifiability trumps citation consistency guidelines. I hope it's also uncontroversial, to state that guidelines are important too, in particular CITEVAR, because it makes it easier for editors to modify and expand an article smoothly and efficiently, if a single citation method is being used, and making it easier for editors to do that ultimately serves our readers, which is the whole point.

So, where does that leave us? I guess what I'd like to say, or rather ask, is, what best serves our readers, here? I can state my bias now: namely that I prefer using {{sfn}} (as you apparently do), but I don't let that get in the way of supporting CITEVAR, and I've reverted users who agree with my bias in order to undo their sfn's and reinstall inline ref's in an article, when that is the consistent pattern. Secondly, if I were more persuaded by your arguments that would be exceptions to CITEVAR—maybe little isolated gardens of sfn's within an article that is mostly inline-cites, say—then I would support you and go along with the idea in the spirit of "verifiability policy trumps MOS guideline", but I don't see that, here. Instead, it feels more like your arguments are mostly in the nature of, "what's easiest for me, or for like-minded editors". Perhaps I'm misunderstanding your motivation; if you can recast your argument about why there should be an exception to CITEVAR in terms of how it helps readers and helps verifiability (or at least, does not hurt it), then I might be more willing to buy your argument. Otherwise, it feels like you are in a WP:ONEAGAINSTMANY situation. Mathglot (talk) 23:55, 8 March 2023 (UTC)

Is an exception required for WWII article to retain their current style? Many here seem to believe that that GA article's style is inconsistent and should be fixed in order to conform to Citing sources guidance. Perhaps some feel they should never have been given a GA and for their next review would have to fix this in order to retain it.

The concrete outcome I'd like to see and which would make WP better is to agree on some new text for citevar giving better clarity that mixed inline and non inline references are not necessarily discouraged if the citation rule is consistently used as appears to be the case in WWII. If that is worded as an exception, of course I will go along with that. Actually I see a lot of common ground here, and honestly I may be a less of a fan of sfns than Mathglot. There are advantages to inline as well as sorted bibliographic sections- As a researcher, I value organized bibliographies in books and for subjects I am familiar with, looking at the references used for a book or WP article pretty quickly tells me the cut of the its jib- Besides verifiability it useful for many other quality purposes- to scan for undue weight or missing significant authorities (whether I agree with them or not) who have not being cited from. But yeah, for verifiability, it is more difficult to hide self authored and dubious content- I have caught a lot of junk that way.

Regarding common ground, of course we all agree about consensus before style changes. But perhaps it may surprise many here that I firmly agree with David's point for example with cases like the WWII article that a citation should not be added to the further reading section for the sole reason of anchoring a new sfn.

Whatever proposed wording change to citevar we can come that addresses this issue would greatly benefit WP. J JMesserly (talk) 01:37, 9 March 2023 (UTC)

See above regarding perfect adherence to highest citation-organization standards not actually being a GA criterion. The World War II article definitely has problems, e.g., the footnotes refer to Bix's Hirohito and the Making of Modern Japan in three different ways. It has both "Notes" and "Citations", with some of the latter more suited to the former. That's the kind of thing that might come up if the article was nominated for Featured status, but I doubt any GA reviewer would find it a make-or-break issue. XOR'easter (talk) 02:56, 9 March 2023 (UTC)

It is progress to have a case example whose style can be taken seriously.

But if everyone else here agrees that WWII's style of mixed inline and non inline references is pathological, then I am a minority of one and mathglot's point about oneagainstmany applies. I am surprized that WWII was given a GA in 2013 if David is right that "avoiding the use of further reading sections for publications that are used as sources actually is required". Because WWII uses them as sources extensively (at that time using harv templates). As my edit history for the last 15+ years here and on Commons bears out, I generally avoid these sorts of disputes like the plague and would rather be improving WP articles.

Quality of WP articles should be an interest of everyone here, so regarding "improvement" of WWII to meet the "highest citation-organization standards", let's set aside the unrelated citation inconsistencies XOr listed that most of us would spot and take care of in a similar manner. In the case of sfns, I am not so sure. What do editors think would best practice be?

• if convert to all inline, to replace sfns with inline duplicates of the citation from the Further reading list? Or would the citation be deleted from Further reading
• If convert all inline <ref>s to sfns, then include fairly non notable citations used a single time alongside those to authoritative texts which are used dozens of times?

From my view both purist approaches would degrade the article's quality and it seems to me that WWII has struck the kind of hybrid balance that would benefit not detract for most if not all of the comprehensive WP articles. I agree with XOR the reference categories could be improved to make them more useful for the reader. J JMesserly (talk) 03:36, 9 March 2023 (UTC)

World War II was not promoted to GA in 2013. It was promoted in 2010. At the time, it followed David Eppstein's style #1: References that are used only in one way (books used only once, web pages, or journal articles) have the reference listed in full in the footnote, but book references where different footnotes refer to different pages or page ranges are listed with a short footnote and the full reference in a later section. I may have missed some glitches when reviewing it today, but it's certainly closer to a consistent standard than what the article has going on now. XOR'easter (talk) 14:20, 9 March 2023 (UTC)

It's beside the point whether it was peer reviewed and retained its GA in 2013, or whether it was first awarded in 2010. In 2010 the article was awarded a GA despite having 35 harv templates used in the style I describe. In 2013 it retained its GA despite having over 70. The question remains- does the citing sources guidance mean that good articles ought to be purist regarding use of sfn, harv, and similar templates supporting non inline references? If so, how do editors suggest that WWII be improved in this regard? I listed the options above, and I think we would all agree that none of them would deliver a higher quality article.

Perhaps folks are getting trapped by the language we use? Note that WWII does not call it a Further Reading section. If some source is cited, it is listed in the citation section using reflist. If it is commonly used as a reference for WWII, it goes in the References section. The principle David brings up remains that people should not add a cite to this section simply to anchor an sfn. The citation information should be there only if it meets the criteria for the section. J JMesserly (talk) 19:25, 9 March 2023 (UTC)

Peer review isn't a process that can take away GA status. That's Good Article reassessment.

Then and now, the World War II article uses zero references in the manner you describe. It has no "Further reading" list. "Further reading" is a different thing from "References"; the Manual of Style is perfectly clear on that, and eliminating accidental redundancy between the two is routine citation maintenance. (Personally, I'd say that merely throwing a book into a list is a lousy way of recommending it, and duplicating a book between "References" and "Further reading" reads more like carelessness than an indication the book is noteworthy. Like "External links", "Further reading" sections are often crufty.) During this discussion MrOllie, David Eppstein, and I have all linked to that same section of the Manual, so I don't see how the term could be ambiguous. XOR'easter (talk) 19:50, 9 March 2023 (UTC)

Protests about the case example do not make the issue go away. The case example uses both inline and non inline references and was independently judged to be a high quality WP article. If it is correct that its references should either be all inline or all non inline, how do editors here propose to fix it? One of the options I listed above, or something else?J JMesserly (talk) 23:11, 9 March 2023 (UTC)

You are the only one arguing that the distinction here is between "all or none inline" and "mixed". The distinction is between consistent and inconsistent. As already has been made clear, it is possible for a style to mix full-reference footnotes and short-ref footnotes while still being consistent, but this does not imply that all such mixes are consistent. The style of your hand-picked example, at the time of its GA pass, was consistent, despite being mixed. The style of your edits to emotion was inconsistent, despite also being mixed. It is invalid to use this example as a justification to whine on and on and on about how unfair it was that your inconsistent formatting was reverted. —David Eppstein (talk) 23:22, 9 March 2023 (UTC)

But we have common ground on consistency David. It would be wonderful, but I am not so sure we have common ground that there can be a mixture of inline and non inline while still being a consistent style. @XOR'easter, do you agree it is consistent to have a mixture of inline and non-inline as in the WWII example? My understanding was that you believe it should be one way or the other due to what the guidelines say. J JMesserly (talk) 23:37, 9 March 2023 (UTC)

Discovering Trappist the monk's {{Ref info}} template above inspired me to wrap it for convenience of use at article Talk pages. As a couple of articles are mentioned above (World War II and Ludwig Wittgenstein; did I miss any?), I've added a {{Ref info banner}} to the Talk pages of those two articles, where you can expand them to see the details. Thanks again, Trappist, and I hope the wrapper encourages wider use of this very informative module. Mathglot (talk) 06:15, 9 March 2023 (UTC)

Thanks for creating a way to add it in a collapsed manner. I've found Ref info to be extremely helpful in working on articles. StarryGrandma (talk) 20:23, 9 March 2023 (UTC)

Your welcome; it was well worth the effort. The current version has a performance weakness (explained at its Talk page) and may time out on articles with hundreds of references, but hopefully that will be addressed in the future. Mathglot (talk) 20:56, 9 March 2023 (UTC) (Update: performance enhancement now in effect, after module update by Trappist. Mathglot (talk) 02:23, 10 March 2023 (UTC)

To summarize the earlier discussion, it was identified that an edit made to the Emotion article ran afoul of multiple guidance principles.

1. that a change to the citation style of an article not be changed without first gaining consensus.
2. that very rare use of non inline references is not sufficient evidence of a mixed inline and non-inline citation style.
3. that whatever mixed inline and non-inline citation style is used, that it be consistent.

• (Terminology note: "Non-inline references" refers to use of short footnotes to a section with an alphabetized list of citations typically using sfn, or harv templates. Inline references means the typical method of citation using the <ref> tag.)

Everyone appears to be ok with all of these points and that the change to the Emotion article should not have been made.

Everyone also appears to be in agreement that David Eppstein provided a helpful example of a consistent style which has both inline and non inline citations: For an article which has multiple citations to different pages of the same source, the citation is placed in a different section, and an sfn or similar non inline reference is made to that citation. The different section must not be the article’s Further Reading section. The World War II article generally follows this style.

<><><> End of summary <><><>

If anyone can't go along with this summary or feel it could be otherwise improved, please feel free to directly edit the summary above. J JMesserly (talk) 20:50, 10 March 2023 (UTC)

I'm not comfortable directly editing a message you left above, as it will invalidate further responses. I object to parts of the above:

• sfn or similar non inline reference... — you may be right here, but in my view, an sfn is "inline" (well, obviously it is, right? that's where it's located); maybe it's not an "inline reference" but an "inline citation", but now things get confusing, because the terms reference, citation, and footnote (and sometimes even source) are confusingly used to mean the same thing. If we are talking about changing wording, then we had better be explicit, use one word throughout, avoid elegant variation, and explicitly list somewhere other terms that mean the same thing (possibly in an WP:EXPLNOTE), or even a whole Terminology subsection, if it deserves it). Somewhere, I hope, there's a page that discusses these terms, and whether we consider {{sfn}} an "inline reference", an "inline citation" or even, "not inline" (that would be bizarre to me, but whatever).
• For an article which has multiple citations to different pages of the same source, the citation is placed in a different section — I cannot agree to this at all. This would be a break with existing practice, and any agreement we came to here, would be WP:LOCALCONSENSUS, and a much higher level of consensus would be required for this imho than what we can achieve in this discussion. There are perfectly adequate methods of dealing with this now, using named references inline with affixed page numbers, so it's absolutely not necessary, and I would say contrary to guidelines, to require multiple citations to different pages of the same item to go to a different section. Big oppose on this part of it.

The text below was interpolated in the middle of an earlier comment, which continues below the box.

The text below is the continuation of the earlier comment by User:Mathglot.

• The different section must not be the article’s Further Reading section. — true enough, as far as it goes, but it must not be in the Infobox, either, or the lead; the point being, if you're going to propose something, isn't it clearer to state what it is, rather than what it isn't? In any case, it may be a moot point, given my previous objection.

There's even another issue which no one has addressed afaict, and I hesitate to raise it because it's a complicating factor and I feel like we're drowning already, or at least, treading water, but you also have multiple citations to the same item which have different quotations (param |quote=) which are applicable to each one. My initial inclination is to be aware of that issue but not to deal with it now, but I wanted to raise it now in case it turns out to be indissolubly linked to the whole, rather than a separate piece that can be dealt with later, as I assume it is. Mathglot (talk) 23:12, 10 March 2023 (UTC)

If David's example was used in the guidance I think I speak for everyone here that it would be clear that it is only an example, and not a requirement for all mixed styles. Does that help? Also, I am ok with dropping the Further Reading admonition. Regarding substitutions of terminology replacing "non inline footnotes" and other terms, I am ok with any other concise terminology that does the job. J JMesserly (talk) J JMesserly (talk) 00:05, 11 March 2023 (UTC)

I've seen the need/desire for multiple quotations from different pages not being supported by the templates as a potential problem, yes. I'm of the mind that this can be addressed using notes (ref/nt), though I've not needed to do this myself. --User:Ceyockey (talk to me) 01:54, 11 March 2023 (UTC)

I'm a little confused what we are trying to accomplish now, so let me summarize where I think we are:

• The OP section title poses this question: "Does CiteVar ban use of sfn templates if they are already used in an article?"

  We've answered this already per discussion, haven't we? Namely it's not about banning, but rather maintaining a consistent style. The guideline notwithstanding, various examples of articles that violate it were named. These are failures to follow guideline, or "violations" if you prefer, but that doesn't invalidate the guideline as the objective to aspire to. I don't see any wording changes that need to be made to § Variation in citation methods (a.k.a. WP:CITEVAR) as it already covers this.

• It seems an edit made to the Emotion article which fails to observe/violates CITEVAR; it should be fixed, but that's a discussion topic for Talk:Emotion, not here. But fwiw, I don't think anybody's objecting to fixing Emotion.
• This subsection title mentions "Next steps", but I'm a little unclear on what you want to do next. I interpret your comments regarding David's example as a wish to change the wording of WP:CITEVAR, although you didn't explicitly say so. Assuming that is your wish, could you recast it as a proposal in explicit terms, such as: Change 'X' to 'Y', or BEFORE: Fooo, AFTER: Barrr. You may find the template {{TextDiff|1=ORIGINAL TEXT|2=PROPOSED TEXT}} ideal for this purpose, but any format that makes your proposal clear works for me.

Did I miss anything as far as "next steps" or anything else still on the table? Mathglot (talk) 01:14, 11 March 2023 (UTC)

I agree with what you stated, so if everyone is in agreement then yes, that is the next step I'd like to explore. The wording could clarify some ambiguities in this regard in the guidance. You appear to have reservations about David's example, but if you have a better example, I am open to using that. J JMesserly (talk) 05:16, 11 March 2023 (UTC)

My only reservation, if you want to call it that, is that I think that WP:CITEVAR is fine as it is, and needs no changes in text, or in examples. I'm open to being persuaded otherwise, but to achieve that someone wishing to change the wording should propose something concrete. Mathglot (talk) 05:25, 11 March 2023 (UTC)

I'm also in the "WP:CITEVAR is fine as it is" camp. Or, if it has problems, this discussion hasn't really served to expose them, and they lie somewhere else. XOR'easter (talk) 05:30, 11 March 2023 (UTC)

It is unclear to me what I as am editor am supposed to do to fix Egyptian Vulture or World War II articles. That's why I asked that very question earlier. They aren't consistent now, and even when WWII has been reviewed it wasn't then. For example books with different page numbers did not appear as non inline citations, and some of the items in the cited sources section were not books or did not have citations to different pages. Yet CiteVar prohibits me from moving anything from non inline to inline or vice versa unless the citation violates a consistency style I can only guess at. They were never agreed to or described for the article nor is there any clue from the Citing sources guidance. Do we expect editors to go cowboy and make something up that seems to fit what is there? At least if some examples were listed I could say for the vulture article- it looks like #4, so that's why I am moving these inline refs to the cited sources section, or for WWII, hey- this is not a book, or there aren't multiple pages cited, so it doesn't follow #1, so that is why I am moving the citation back inline. J JMesserly (talk) 06:06, 11 March 2023 (UTC)

I think WP:AIN'T BROKE applies here. Your comment appears to assume the existence of some policy or guideline about "books with different page numbers [that do] not appear as non inline citations", but there is no such guideline afaict, so nothing there needs correction on that basis. Likewise, it may be that "some of the items in the cited sources section were not books or did not have citations to different pages", but that does not run contrary to any guideline or policy I'm familiar with, and therefore does not need to be corrected either. It could well be that Egyptian Vulture does need adjustment for citation consistency (I haven't looked), but if so, it would have to be based on actual guidance.

But it does feel to me like we're going round in circles, alternately seeking how to interpret the current guideline, or how to change the wording to something which would clarify any points of confusion in the current guideline. I'm losing the sense of any clear forward momentum, and I think I've contributed all I can of a positive nature for the time being and would start to repeat myself if I stayed. Rather than do that, I'm going to unsubscribe for now, and let others have their say and bring fresh eyes and fresh opinions to bear on the topic. I wish you luck, and perhaps I'll pop back in to see how things are going. Best, Mathglot (talk) 07:54, 11 March 2023 (UTC)

Not broken? The guidelines prohibit moving citations from inline to non inline form if such a move violates the style of the article. To be concrete, I gave an example of an article- WWII, and the question was- what is the style being used that will allow us to know how to "fix" its inconsistencies? I believe XOR was of the view that there was no consistent style for the mixture, and David proposed a rule that might be used in the future to make it consistent, but in fact it never adhered to that proposed rule. So should am editor faithfully attempting to follow the citing sources consistency guideline simply make up the best rule they can think of that seems to fit the mixture of inline and non inline citations they see? Or should they go to the talk page and solicit a consensus on how the inconsistency should be handled? Or is it best practice for editors to declare no consistent style is in place, so they then have free hand to establish a style of their choosing?

Sure, we can leave the guidelines vague, but I don't see that this moves us towards mixed articles ever being consistent. J JMesserly (talk) 10:06, 11 March 2023 (UTC)

There's not going to be a one-size-fits-all procedure. If all the references in an article are {{sfn}}s that point to items in a "Works cited" list, and someone adds a new one that's just a footnote containing a {{cite web}}, then the new reference should be converted to the style of all the others. If an article has been sitting around a while since its last top-to-bottom overhaul, then maybe it has mostly {{sfn}}s but 10–20% other styles. Then imposing a single style would be a bigger job, and it might be fair to ask on the Talk page first: "It looks like most of the citations are style X, but 10% are in style Y, and there isn't any obvious reason for the difference. I think we should convert them all to style X." XOR'easter (talk) 13:34, 11 March 2023 (UTC)

This dispute has been very educational for me, so thanks, everyone. It has made me realize I did one of the sources wrong on the Ängelholm UFO memorial which I have now fixed so that it no longer points to Further Reading.

When using both {{sfn}} and ref tags together, is it acceptable to have only a few sources (instruction manuals and changelogs in [case]) in a bibliography for sfn? Or do the sources need to go something like first use the full source in ref tags, and then later refer back to it with {{sfn}}? Rjjiii (talk) 04:54, 12 March 2023 (UTC)

I think David's answer for acceptable mixtures is the one most agreeable to participants here. However I am not so sure, because the sfns to the bibliography will produces visual inconsistency- the reader will see the full citation when mouseover most footnotes but not those using a short footnote template. If different page numbers is the motivation for the sfn, then an alternative is to use Template:RP after a ref name= link. This has the additional advantage of full Visual Editor support, making them less onerous to add than sfns. See Parrot for an example of rp use. However, RP is limited in that there is nothing like sfn's loc field which can be use for quotes, annotations and links. For such links, Parrot happens to have an example of using sfns which will display links to pages on google books. This is kind of cool, but editors such as XOR thinks such references are unstable and should be discouraged. I don't do them myself, but I have not observed any instability with google page links, so I don't know that I agree.

I am a big fan of the principle "If it aint broke dont fix it", but Parrot is yet another example of inconsistent citing styles. The problem is rampant.J JMesserly (talk) 21:26, 12 March 2023 (UTC)

The template documentation at {{Rp}} explains how to use it to display quotations. MrOllie (talk) 21:41, 12 March 2023 (UTC)

I've observed that instability, and it's mentioned in places like the "Citing sources" guideline and this essay. I am not opposed to Google Books links in general, but they can't be depended upon. If there's a better alternative, use it; if providing a Google Books link doesn't actually help more than providing the OCLC and/or ISBN, we might as well omit it. In our discussion at Talk:Bird_vocalization, I was making the point that a link shouldn't be used in the url= field of {{cite book}} if it doesn't actually point to a copy of the book. (The case in question involved a web page with a description of and quotation from a book, but no way to purchase it, locate it in a library, or read more of it.) XOR'easter (talk) 00:02, 13 March 2023 (UTC)

@J JMesserly: Thanks! the reader will see the full citation when mouseover most footnotes but not those using a short footnote template Yeah, that makes sense; I'll use a mix of mostly {{rp}} and a few redundant references where I have links plus quotes. The rp template seems like the default way to handle multiple-use citations, but it also feels pretty limited.

@MrOllie: Thanks for the tip. It's cool that rp can do this but it's really lacking compared to {{sfn}} and {{harv}} because the quotation is unavailable in mobile, higher threshholds of dexterity/eyesight are necessary to access the quotation, and I would expect it's not clear to non-editors how to access the quote. Again thanks for the heads up. Rjjiii (talk) 00:52, 13 March 2023 (UTC)

The rp template seems like the default way to handle multiple-use citations. I'm not so sure. In the Birds Wikiproject articles, sfn is used in around 300 articles. RP is used in a little more than one hundred, and frequently the article has it used a single time. Personally, I am ok with either style. I like sorted sources lists you get with harv and sfn, but besides the VE ease of use and consistency with full citation popups, the rp method does have value for raising citation quality. Doing rps to a single ref will give the citation counts in the reference section- an indicator of overreliance / underuse of particular common references for a subject. This is something you don't get using an sfns / harv style. J JMesserly (talk) 01:32, 13 March 2023 (UTC)

User:J JMesserly, you state above in reference to your two example articles, which I just had a look at, "some of the items in the cited sources section were not books or did not have citations to different pages." Can you explain why this is a problem? I've been citing journal articles, books with a single cited page, and books with multiple cited pages identically since forever, and I'm wondering how this is not best practice. Folly Mox (talk) 22:56, 12 March 2023 (UTC)

Folly Mox, sorry for the confusion. I don't think it is in general a problem at all- that is, I don't think mixed articles need to follow that particular rule, and to be fair I'm not sure the author of that rule intended that it applied to WWII.

Some context was omitted. XOR early had asserted that WWII "was promoted in 2010. At the time, it followed David Eppstein's style #1:" (This is a rule requiring that a source have different pages cited, and the source be a book). To make the point that such rules ensuring rigid consistency are not all that crucial and are problematic to enforce, I pointed out that this GA article in fact did not follow Eppstein's rule, because 1) at the time (see xor's link above) there was large PDF that was not a book so should have not been in the cited sources section, and 2) there were some book citations in the reflist section which cited different pages and so should have been moved to the "cited sources" section.

Which doesn't make the rule a bad one. David rule #1 is a good example of a rule for consistency because it happens to conform to how a large number of mixed articles are using sfns. I was thinking that if CiteVar included an example of a consistent rule that this would be a good thing, but there appears to be no consensus for that.

I can imagine that the regular contributors of WWII article might choose some other rule such as one that had to do with the quality of the source being cited. OTOH Editors there more likely may think it is a low priority issue- That since being "consistent" in this way is not important enough that it is part of a GA evaluation, perhaps it's not worth going through the hassle of establishing by consensus what the article's style should be regarding sfns.J JMesserly (talk) 23:31, 12 March 2023 (UTC)

Interested editors are invited to look at WT:ISBN#Hyphens in ISBNs. – S. Rich (talk) 02:32, 16 April 2023 (UTC)

Any tool which automatically removes redundant spaces from citations? Shahid • ^Talk2me 15:48, 23 April 2023 (UTC)

You could use regular expressions to do this. Eg (?<=\w) (?=\|) and (?<=\|) (?=\w) to identify the relevant spaces; if you're altering citations in the running text, this is more difficult. You may need to copy the text into an external editor like Sublime Text or Atom first. Ifly6 (talk) 16:50, 23 April 2023 (UTC)

@Ifly6: What do you mean by these expressions? How do you use them? Shahid • ^Talk2me 17:17, 23 April 2023 (UTC)

Regular expressions allow you to match patterns in text. See https://www.regex101.com for an explanation of them. I use forward and backward lookaheads for pipes and word characters. You need a regex engine to match and replace. Sublime Text and Atom are both text editors that have such an engine built in; the 2017 visual editor does too but (as I noted in a bug report recently) it doesn't work. Ifly6 (talk) 17:58, 23 April 2023 (UTC)

How should we handle the dating of identical reprints from the same publisher? There was a previous discussion on this, it seems, but there were few comments on it. Ifly6 (talk) 17:37, 10 April 2023 (UTC)

When I've run into this situation, my solution was to follow the cite template with a parenthetical note like "(pagination identical to X edition)". That way it's possible to use the version on hand while making it clear to the reader there are options for verification. I'm not 1000% clear on what problem it is you're trying to solve here, so apologies if my suggestion misses the point. Folly Mox (talk) 19:11, 10 April 2023 (UTC)

There are multiple reprints for the Cambridge Ancient History 2nd ed vol 8, in 1989 and 2006 at least. The question is what year to use. If you look in the classical scholarship this is always 1989, date of publication because the 2006 reprint is identical. Sfn and Harvnb anchors accept only one year. I can't say I see the sense in citing an identical reprint of a previous edition, especially when – in the citation – it makes the source look newer than it is. Ifly6 (talk) 21:25, 10 April 2023 (UTC)

Yeah, I agree with your point that if you're leaning on interpretative scholarship it's probably best to link the ideas to the time they were put into print. The state of the field from which the cited information arose is better reflected by the earlier publication date. Folly Mox (talk) 00:11, 11 April 2023 (UTC)

Wouldn't the |edition=2 field tell you that your 1989 copy is the same as the other editor's 2006 copy? If they are the same edition then they would be for practical intents and purposes the same.  Stepho  talk  00:39, 11 April 2023 (UTC)

I'm confused: are you suggesting that my copy printed in 1989 would have a note in it of a future reprint in 2006? Or am I just misinterpreting your comment? Ifly6 (talk) 00:57, 11 April 2023 (UTC)

What I mean is that the reference in the article should have |edition=2 in the reference. When the other editor looks at his copy he sees that he has edition 2 and that the reference matches his copy. When you look at your copy you also see that you have edition 2 and that the references matches your copy. The actual print date can therefore be ignored.  Stepho  talk  01:08, 11 April 2023 (UTC)

If that's the case, then why should we pick dates which make books appear newer and more up-to-date than they actually are? Ifly6 (talk) 01:50, 11 April 2023 (UTC)

Indeed, if you have the edition number then a case could be made for not bothering to list the print year.

But if we are going to put the print year in (as most editors prefer to do), then you should start with the year of the book you have in your hand. If you are able to reliably figure out the year of the first print (often listed in a table of print years in the front matter) then it might be nice to list that year instead (as per your argument). In either case, a reader with a book with different print year but the same edition number would know that their copy will match what the other editor had, even if the years do not match exactly.  Stepho  talk  02:40, 11 April 2023 (UTC)

Do you think we should be listing the first printing of an identical edition or the year of the printing in hand? Ifly6 (talk) 17:05, 11 April 2023 (UTC)

I follow the advice given in The Chicago Manual of Style (17th ed.) paragraphs 14.142 and 14.143.

For books, only the year, not the month or day, is included in the publication date. The date is found on the title page, or more commonly, on the copyright page. It is usually the same as the copyright date....

The publication date must not be confused with the date of a subsequent printing or a renewal of copyright. Such statements on the copyright page as "53rd impression" or "Copyright renewed 1980" should be disregarded.

Jc3s5h (talk) 17:19, 11 April 2023 (UTC)

I presume, given that you were the one who made the suggestion to clarify guidance in the discussion I noted at the top of this query, that you still would have that guidance explicitly clarified? (I would hold the same view; it is what academics do.) Ifly6 (talk) 17:54, 11 April 2023 (UTC)

Since my previous comments were about 8 years ago, I don't know if I would say the same thing now, because I don't recall how much the guideline has changed from then and now. If you intend to change the guideline, I suggest you give before and after versions on the talk page. Jc3s5h (talk) 22:35, 11 April 2023 (UTC)

Given Jc3s5h's previous comment, I would propose rewording the section on dates and reprints:

Dates and reprints of older publications

Date a book that is identically reprinted or printed-on-demand to the first date in which the edition became available. E.g. if an edition of a book was first released in 2005 with an identical reprinting in 2007, date it to 2005. If substantive changes were made in a reprint, sometimes marked on the verso with "Reprinted with corrections", note the edition and append the corrected reprint year to it (e.g. "1st ed. reprinted with corrections 2005").

Editors should be aware that older sources (especially those in the public domain) are sometimes reprinted republished with modern publication dates; treat these as new publications. When this occurs and the citation style being used requires it, cite both the new and original publication dates both the original's publication date, as well as the date of re-publication, e.g.:

• Darwin, Charles (1964) [1859]. On the Origin of Species (facsimile of 1st ed.). Harvard University Press.

This is done automatically in the {{citation}} and {{cite book}} templates when you use the |orig-date= parameter.

Alternately, information about the reprint can be appended as a textual note:

• Boole, George (1854). An Investigation of the Laws of Thought on Which Are Founded the Mathematical Theories of Logic and Probabilities. Macmillan. Reprinted with corrections, Dover Publications, New York, NY, 1958.

Comments welcome. My additions in green and removals struck through. Ifly6 (talk) 23:23, 11 April 2023 (UTC)

Your proposal sounds good to me.  Stepho  talk  23:27, 12 April 2023 (UTC)

Thanks. I also made a few changes to my proposal (largely cosmetic about wording). Ifly6 (talk) 01:32, 13 April 2023 (UTC). Added the following E.g., if an edition of a book was first released in 2005 with an identical reprinting in 2007, date it to 2005 to be more explicit. Ifly6 (talk) 02:31, 13 April 2023 (UTC)

Added and changed some wording that now characterises the modern reissues of historic books as "reissues" rather than "reprint"s along with text specifying that these should be treated a new editions. Ifly6 (talk) 13:36, 13 April 2023 (UTC)

I'll reiterate that I agree when possible we should be citing the publication date of the information rather than the manufacturing date of a particular book. Does this place any additional burden on editors with later printings to confirm their edition is identical to the first printing? Will it necessarily be obvious enough from the copyright page? Folly Mox (talk) 17:09, 13 April 2023 (UTC)

Most of the sources that I've worked with are academic works which are pretty clear on the copyright page as to when they were first printed. However, more generally, when it comes to this sort of thing, I don't see much of a difference between are-reprints-really-identical and is-the-Google-Books-version-really-identical. We largely, and most of the time reasonably, trust that they are. Ifly6 (talk) 01:13, 14 April 2023 (UTC)

@Jc3s5h: Do you have any feedback on this proposal? — Preceding unsigned comment added by Ifly6 (talk • contribs) 14:58, 15 April 2023 (UTC)

Because the original discussion woke back up, I should disclose that my initial question emerged from this thread. Ifly6 (talk) 20:33, 15 April 2023 (UTC)

@Stepho-wrs, Folly Mox, and Jc3s5h: Do you know how we actually implement changes on these pages? Do I just edit it in? Or is there some kind of informal process that needs triggering? Ifly6 (talk) 14:21, 17 April 2023 (UTC)

Once there is consensus on what the change should be, Ifly6 could certainly be the one to perform the edit. A best practice would be to mention this discussion in the edit summary.

I don't think I would use the phrase "verso page". Someone in need of this guidance may not know what a verso page is.

Also, I wouldn't treat a book that has been reprinted with corrections as a separate edition, because the book will be marked as whatever the original printing was. By failing to state what the publisher considers to be the edition number, readers might be confused about whether they have the same source in their hands that the Wikipedia editor used while writing the article. I would, where the edition is given, write something like "1s ed. (reprinted with corrections 2003)". Jc3s5h (talk) 14:45, 17 April 2023 (UTC)

The reference to verso is already live on the MOS page, I used it to be consistent with existing terminology, though "copyright page" is probably more clear. I've made the clarifications suggested. Ifly6 (talk) 18:48, 17 April 2023 (UTC)

Again made a change or two to the suggestion above, taking wording notes from Waddingham New Hart's rules (2d ed, 2014) p 6:

An edition is a version of a book at its first publication and at every following publication for which more than minor changes are made: a book goes into a new edition when it is revised, enlarged, abridged, published in a new format, or published in a different binding. A new edition requires a new ISBN (see below).

A reprint or impression is a republication of a book for which no corrections or only minor corrections are made. The publishing history usually distinguishes between these two states, describing them as 'reprinted' and 'reprinted with corrections'. The publishing history usually details the issuing of multiple reprints in a single year: Reprinted 2004 (twice).

All bolding in original. Are there any objections to putting effecting the noted changes? Ifly6 (talk) 21:15, 22 April 2023 (UTC)

Do not object. Folly Mox (talk) 23:42, 22 April 2023 (UTC)

Does anyone think I ought to wait longer or just push the changes to live? Ifly6 (talk) 21:38, 25 April 2023 (UTC)

Good morning, I have been asked to write an article about a food company. I have two independent sources to cite in this article. Is that enough sources or do I need a third? TaraB1216 (talk) 08:35, 16 June 2023 (UTC)

It depends but there is no rule. There are countless articles with one source and even some with none. I think a good way to view it is to consider the size and complexity of the article and then relate that to the number of sources needed. Make sure you can get truly independent RSSs, not just the company's website. Just my thoughts. Roger 8 Roger (talk) 10:27, 17 June 2023 (UTC)

Unless those sources are very in-depth (longform magazine articles, books) I wouldn't say it's enough. If you can only find two sources that implies minimal coverage and an article failing the GNG. Der Wohltemperierte Fuchs ^talk 11:46, 17 June 2023 (UTC)

@TaraB1216 Having made your WP:PAID declaration, your next step could be to specify the sources you plan to use here in this thread, for experienced editors to comment. A quick Google of 4CFoods led me to a brief listings in Bloomberg and Crunchbase, together certainly not enough to meet our notability guidance. You might like to read WP:BOSS. Mike Turnbull (talk) 11:59, 17 June 2023 (UTC)

Hello, I have been asked to create a Wikipedia article about 4C Foods. I have been given these sources to write the article. Are these acceptable sources? It would be great if someone could let me know. Thank you! Tara

Griffin's Modern Grocer, October 1999, Pages 2-4, 4C Foods Corporation Special Section

Food Trade News, November 2004, Volume 58, Number 11, "4C Foods: Combination of Tradition and Innovation Spells Success" TaraB1216 (talk) 14:26, 17 June 2023 (UTC)

@TaraB1216 The problem with these sources is that they are not WP:independent. The first has clearly been written by, or on behalf of, the company itself and the second is based mainly on interviews and a visit by a trade journalist. The stories are arguably interesting but not sufficient to make the company notable as defined here. They are the sort of content I would expect to see on the company's own website as "history", not in an encyclopaedia. If you had some truly independent sourcing, some of the details in these sources could be used as described at WP:ABOUTSELF. Mike Turnbull (talk) 10:33, 19 June 2023 (UTC)

Across at the Teahouse, I asked: if there's an emoji in a source's headline, should it be left out of a citation title (unless essential to understanding the headline) or included? An example (the context being celebrity gossip) would be:

• "Wait, what? Love Island's Leanne Amaning and Mike Boateng Reunite 😱". Heat. 2021. Retrieved 2023-06-23.

rtrb carried out some useful detective work and answered as follows:

"Looking at WP:CITESTYLE, it implies that deferring to style guides like the MLA is acceptable, and the MLA provides something pretty close to what you are describing: "Emojis do not need to be included in the titles of tweets. If the emoji is important to the meaning of the tweet, you should describe the emoji in your prose. If the tweet’s text is composed solely of the emoji itself, its title in the works-cited list should be a generic description of the emoji."^[1] WP:REFNAME (for footnotes and repeated citations) says that the short names of those references should generally only contain ASCII characters, which would disallow emojis there (although from your example it would be unnecessary anyway)."

If this is not already covered somewhere at Wikipedia, could this be worked more directly into Wikipedia:Citing sources or another relevant page?

Many thanks. Esowteric + Talk + Breadcrumbs 16:33, 26 June 2023 (UTC)

I don't think MLA's guidance reflects broad consensus. Eg APA seems to say to retain emojis and hashtags where possible.^[2] Emoji are taken in evidence in American courts, which indicates that they should be retained. Nothing yet, however, is in the Bluebook. Ifly6 (talk) 17:34, 26 June 2023 (UTC)

@Esowteric, I suggest doing whatever you (and any other editors at that article) think is best (or easiest). It just isn't that important. WhatamIdoing (talk) 19:20, 27 June 2023 (UTC)

In some cases including the emoji may be necessary to make sense of the citation title: it cannot just be omitted without changing the meaning, because it is used as part of the title content rather than just as punctuation. An old example: boyd, danah (2008). "Why youth ♥ social network sites: The role of networked publics in teenage social life". In Buckingham, David (ed.). Youth, Identity, and Digital Media. MIT Press. pp. 119–142. In this case you could work around this by replacing the heart symbol by the word heart (possibly in parentheses: "Why youth (heart) social network sites"), and some citations to this work do that. But in general we should use citations that accurately represent the source material, so if the source puts emoji in the title I think we should do the same rather than textual workarounds. —David Eppstein (talk) 21:22, 27 June 2023 (UTC)

I think these are pretty reasonable explanations and arguments in favor of retaining emojis in references. Even though this is a niche problem to have, it might be worth adding some words to the article about this, probably in WP:HOWCITE under the "What information to include" section. Something like this maybe:

"When citing, all information related to the source should be kept to be as accurate as possible. Titles should not be abbreviated or have text removed from them, and if a title has non-Latin characters, emojis, or any other special symbol, then that text should be retained within the citation. However, when using a source more than once the name of the reference can and should be shortened, and strictly ASCII characters are preferred."

I doubt it would be necessary but someone could also add an example using the "size of the sun" text. rtrb (talk) 16:32, 29 June 2023 (UTC)

I'm not entirely sure as to whether it's necessary to add explicit guidance on emojis. Just say Titles should be entered verbatim. Ifly6 (talk) 22:19, 29 June 2023 (UTC)

Thanks a lot: what you've each written makes a lot of sense. Esowteric + Talk + Breadcrumbs 06:53, 28 June 2023 (UTC)

The description of linking to a page in a PDF is misleading. In general, there are two types of page numbers in a PDF: the page numbers rendered within the document and the position of the page within the PDF. Often these are different, either because the document doesn't start numbering at 1, the document has independent numbering of prefatory material and normal text, or the document has hierarchical page numbers. The page= fragment uses the PDF page number, not the document page number.

Thus, in the citation^[1] {{cite book|title = Assembler Language Programming for IBM System z ™ Servers - Version 2.00|author = John R. Ehrman|date = February 2016|edition = Second|quote = Some people call it "BAL" — meaning “Basic Assembler Language” — but the language is not basic(nor is it BASIC) except in the sense that it can be fundamental to understanding the System z processor's operations.|url = http://zseries.marist.edu/enterprisesystemseducation/assemblerlanguageresources/Assembler.V2.alntext%20V2.00.pdf%7Cpublisher = IBM Silicon Valley Lab|access-date = July 12, 2023}}, the number in the document is for but the fragment in the URL is page=42, the position of page 4 withinn the PDF file. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 11:24, 12 July 2023 (UTC)

I always use whatever page number is displayed on the page, same as if I'm citing an archived book source where "page 1" in the reader is a picture of the front cover. Not all PDF readers even display which page of the file you're looking at (including many browser-based readers), but the printed page number is always visible. I don't think vagaries of file encoding or storage should play a role in which page number to cite. Folly Mox (talk) 11:42, 12 July 2023 (UTC)

PDFs are digital representations of print documents. If they have hard-coded page numbers, use those, the same way that if you were reading a book you would flip to where the book says page 4 is rather than count, inclusive of the cover, the number of times you had to flip a piece of paper. Ifly6 (talk) 12:24, 12 July 2023 (UTC)

Towards the end of the section on linking to Google Books, we find this:

When the page number is a Roman numeral, commonly seen at the beginning of books, the URL looks like this for page xvii (Roman numeral 17) of the same book:

     https://books.google.com/books?id=kvpby7HtAe0C&pg=PR17

The &pg=PR17 indicates "page, Roman, 17", in contrast to the &pg=PA18, "page, Arabic, 18" the URL given earlier.

You can also link to a tipped-in page, such as an unnumbered page of images between two regular pages. (If the page contains an image that is protected by copyright, it will be replaced by a tiny notice saying "copyrighted image".) The URL for eleventh tipped-in page inserted after page 304 of The Selected Papers of Elizabeth Cady Stanton and Susan B. Anthony, looks like this:

     https://books.google.com/books?id=dBs4CO1DsF4C&pg=PA304-IA11

The &pg=PA304-IA11 can be interpreted as "page, Arabic, 304; inserted after: 11".

Note that some templates properly support links only in parameters specifically designed to hold URLs like |url= and |archive-url= and that placing links in other parameters may not link properly or will cause mangled COinS metadata output. However, the |page= and |pages= parameters of all Citation Style 1/Citation Style 2 citation templates, the family of {{sfn}}- and {{harv}}-style templates, as well as {{r}}, {{rp}} and {{ran}} are designed to be safe in this regard as well.

Wikipedia DOI and Google Books Citation Maker or Citer may be helpful.

Users may also link the quotation on Google Books to individual titles, via a short permalink which ends with their related ISBN, OCLC or LCCN numerical code, e.g.: http://books.google.com/books?vid=ISBN0521349931, a permalink to the Google book with the ISBN code 0521349931.

For further details, you may see How-to explanation on support.google.com.

There is no identifiable reason I can find for the Note that some templates... paragraph to be indented. I suspect the idea was that it be read as some sort of admonishment / "note box" (simply based on the fact that it starts with Note...), however that effect is nowhere apparent. Instead, it reads as being randomly indented for no reason.

(It's also indented using a leading :, talk-page-style, rather than using <blockquote>...</blockquote> or some other semantically-correct method. Just another strike against it.)

The only things holding me back from simply being bold and fixing it are: (a) it's part of a Guideline, and (b) it has been like that for a very long time. (The paragraph started off much smaller, but has grown in length from its relatively meager beginnings.) Still, unless anyone objects and can provide a justification for the indent, I propose to simply remove the leading colon and format the text flush to the left margin, in line with the paragraphs immediately preceding and following. FeRDNYC (talk) 10:00, 20 July 2023 (UTC)

The tool returns a 404 when you attempt to reach it

https://alyw234237.github.io/wiki-doi-gbooks-citation-maker/

Anyone know if we can reach out to the person who ran it or at least copy the repository and get it up with a different Github account or on its own domain?--12.231.138.10 (talk) 02:39, 19 June 2023 (UTC)

It appears that that repository was created (or at least offered) as an alternative to http://reftag.appspot.com/, when that went offline ~ 2 years ago. However, it also appears that GitHub user alyw234237 has completely deleted their GitHub account and all repositories associated with it, and a search of GitHub's public repositories does not turn up any other repositories with that name. It's not looking too promising. FeRDNYC (talk) 10:13, 20 July 2023 (UTC)

So, general references is sort of an excuse for editors to link dump a bunch of bare URLs to avoid having to specifically do in-line citations? I've removed a handful of bare URLs and an editor objected and said they shouldn't have been removed citing "general references". Graywalls (talk) 02:56, 8 July 2023 (UTC)

This is where it happened: Special:Diff/1164121567. I removed them originally under WP:ELMINOFFICIAL. Graywalls (talk) 05:01, 8 July 2023 (UTC)

Wikipedia:Village pump (idea lab)/Archive 49#Deprecate future use of general references? is the most recent conversation I can remember about it. I think there's discretion involved. The user in this case has been here since 2006 and has 60,000 edits and should definitely do better. Also it's not like they're combining information from three book-length biographies: they're citing news stories. It shouldn't be too onerous of a burden for them to use inline citations, or like format their references. Folly Mox (talk) 03:21, 8 July 2023 (UTC)

Bare URLs feel like they're already de facto deprecated. If they aren't, I think they ought to be and there should be no objection to turning bare URLs into full citations. As to general references, I had read verifiability and it discusses in-line citations only as the way to meet it (ie it [verifiability] is satisfied by providing an inline citation to a reliable source that directly supports the contribution); wouldn't such vague references violate that policy? Ifly6 (talk) 04:39, 8 July 2023 (UTC)

WP:MINREF (an information page) has Our sourcing policies do not require an inline citation for any other type of material.... Technically, if an article contains none of these four types of material, then it is not required by any policy to name any sources at all, either as inline citations or as general references. For all other types of material, the policies require only that it be possible for a motivated, educated person to find published, reliable sources that support the material.... The workaround, of course, is to tag it {{cn}}. That forces the requirement for an inline cite. Folly Mox (talk) 05:08, 8 July 2023 (UTC)

The section on challenging reminds me of a line in Yes Minister about discrediting studies. Paraphrasing: [If you want to discredit a study] say the results of the study have been questioned. What if they haven't? Question them! Ifly6 (talk) 15:22, 8 July 2023 (UTC)

@Folly Mox:, how do you address the issue of editors using a farm of any URLs that even contain the mention of subject under creative heading when what they're really wanting to do is the common public relations desire to do "in the press" section? Graywalls (talk) 05:12, 8 July 2023 (UTC)

The general concept that writers should avoid plagiarism means that our editors should acknowledge in some way the sources of ideas that we put into an article. If the original editor chose to do that with general references, that may not be ideal, but it's better than nothing. If we remove the citation but don't remove the ideas that came from the general references, we commit plagiarism. Jc3s5h (talk) 11:36, 8 July 2023 (UTC)

User:Graywalls, that's a good question. If the source doesn't support any prose in the article, and doesn't meet SIGCOV, it should probably be removed as cruft / spam / trivia. Also I want to clarify that I do have some concerns with how lenient our current sourcing policy is. All the guidance we give new editors tells them to use inline citations, and AfC regularly declines drafts that don't. I'm not sure WP:MINREF would stand up to an RfC, but see Wikipedia:Village pump (policy)/Archive 176#RFC: change "verifiable" to "verified". Folly Mox (talk) 12:13, 8 July 2023 (UTC)

Harvard_Design_Magazine#Reviews_and_Mentions A section like this is a pretty good example of how marketing and public relations could resort to "general reference" rationalization to dump links to showcase places where they're mentioned. Graywalls (talk) 23:43, 9 July 2023 (UTC)

Who cares if a list of sources is a "common public relations desire"? The article you linked has zero inline citations. When someone adds half a dozen sources, even if they're badly formatted and half of them are Wikipedia:Interviews, that's what we have historically called "improving the article".

A long time ago, our approach to Wikipedia:Conflict of interest rules required an actual conflict of interests, as understood by the literal words. We don't follow that model any longer, but 15 years ago, we understood that a COI was an incompatibility between the aim of Wikipedia, which is to produce a neutral, reliably sourced encyclopedia, and the aims of an individual editor. If there was no "incompatibility", then there was no COI as we (oddly) defined it. We believed then that it was possible for someone to advance Wikipedia's interests (e.g., by adding sources to an unsourced article) while simultaneously advancing personal interests (e.g., by adding sources to an article about the company that employs you).

We have since adopted a more typical corporate notion of COI, but I suggest that when it comes to unsourced articles, that any sources being added is better than no sources being added, even if we could imagine that the subject's PR department might be very happy about the sources that were added. PR-department-approved sources are still sources, and even a single weak source can protect us from WP:Hoaxes. WhatamIdoing (talk) 23:50, 20 July 2023 (UTC)

Can the author of a book add content to a Wikipedia article or correct mistakes based on a book, article or working paper of his own authorship? Can that author cite that source? Can this be considered self promotion? Do we care as long as it improves the content of Wikipedia? Rodolfoaoviedoh (talk) 23:36, 15 July 2023 (UTC)

This is enough of a problem that we have a standard warning template about it, {{uw-refspam}}. In practice this is allowed if it isn't excessive. If you cite yourself rarely, no one will care. If most articles you edit end up with your name in the reference section, that's going to cause a problem. MrOllie (talk) 01:36, 16 July 2023 (UTC)

"Working papers" or preprints are more likely to be seen as problematic than published books and articles; they generally do not meet our requirements for reliable sources.

Also, you should mention the conflict of interest in an edit summary, and it would be a good idea to avoid reinstating the edit if anyone objects and undoes it. (Instead, you could start a discussion on the article talk page on whether the reference is relevant.) —David Eppstein (talk) 02:18, 16 July 2023 (UTC)

The official guideline permitting an limited amount of self-citation is WP:CITESELF.

@Rodolfoaoviedoh, if editors have misrepresented something you've written, please please please let us know. Either fix the article yourself, or post a link (here's fine; we'll find someone who can help, or Wikipedia talk:WikiProject Mathematics is a good option, if it's about math). WhatamIdoing (talk) 23:56, 20 July 2023 (UTC)

Replace:

• Alternatively, use one of the templates listed at the disambiguation page Template:Multiple references

With:

• Alternatively, use one of the following templates: {{sfnm}}, {{sfnmp}}, {{multiref}}, or {{multiref2}}.

Reason:

• Template:Multiple references only lists a handful of templates and is only linked from this page.

Rjjiii (talk) 20:10, 25 July 2023 (UTC)

Oh, this has been rewritten. I have a question about though some of them have unresolved display issues on mobile devices and may eventually merge. Which templates do both parts of that statement apply to? Which templates have display issues on mobile? And which templates will merge?Rjjiii (talk) 02:55, 29 July 2023 (UTC)

In the first paragraph of the Citation style we find a sentence which I think should be changed by striking out some styles, thus:

A number of citation styles exist including those described in the Wikipedia articles for Citation, APA style, ASA style, MLA style, The Chicago Manual of Style, Author-date referencing, the Vancouver system and Bluebook.

I suggest this because we have deprecated parenthetical referencing, and the struck-out citation styles only allow parenthetical referencing. (Chicago Manual of Style allows either footnotes or parenthetical referencing, so the parts about footnotes would still apply.) Jc3s5h (talk) 10:30, 24 July 2023 (UTC)

I think that's reasonable. An exhaustive list of citation styles is not necessary. Ifly6 (talk) 21:07, 24 July 2023 (UTC)

Author-date referencing is still widely used, and not deprecated, within footnotes. What is deprecated is using it as a style of referencing in inline article text, instead of footnotes. We should not alter our guidelines in a way that would be easily misinterpreted as forbidding author-date referencing altogether. It is not forbidden. It is not even deprecated. —David Eppstein (talk) 21:56, 24 July 2023 (UTC)

I see the usefulness of author-date citations in footnotes. But I'm not aware of any published style guide that recommends footnotes in article text and author-date when a citation is needed in a footnote. So I don't think it's helpful to send readers to articles about styles that can't be fully applied to Wikipedia articles. If we want to suggest author-date citations inside footnotes, perhaps we should add something to the guideline saying so. Jc3s5h (talk) 23:11, 24 July 2023 (UTC)

Perhaps you should read this guideline, and specifically its section WP:CITESHORT, before falsely claiming that our guideline does not already recommend that style. —David Eppstein (talk) 00:11, 25 July 2023 (UTC)

I think @Jc3s5h has a point, but I'm not sure that we want to take it this far.

@David Eppstein, here are two quoted examples for the APA style guide, in the section on "In-Text Citations":

• Falsely balanced news coverage can distort the public’s perception of expert consensus on an issue (Koehler, 2016).
• Koehler (2016) noted the dangers of falsely balanced news coverage.

(Another section explains how to write out the full description of the source for placement in the ==References== section; for right now, we are only concerned with the bit that they use as an inline citation.)

The question is: Can you see any way to simultaneously comply with the RFC/deprecation of parenthetical citations and also comply completely with the APA's in-text citation style, which requires the use of parenthetical citations? I can't. WhatamIdoing (talk) 03:30, 29 July 2023 (UTC)

Sure. Use the second format, add a footnote also citing the same source ({{sfn}}):

Koehler (2016) noted the dangers of falsely balanced news coverage.^[1]

1. ^ Koehler 2016

and tell busybody editors who care about such things that the footnote is the citation and the in-text parenthetical thing is just article text describing the author and date of a work, not a citation. They have different purposes: the in-text part is article content telling readers about the history of who noted something, and the footnote provides verifiability to those seeking to verify article content. Only the actual citation in the footnote is covered by the deprecation, because the deprecation was in a discussion about references, not about the manual of style for article text. If someone complains saying the same thing twice in two slightly-different formats looks stupid, tell them that this was a predictable and predicted consequence of the deprecation RFC and that they should redo the RFC while pointing out the problems that it has continued to cause. —David Eppstein (talk) 04:24, 29 July 2023 (UTC)

And if you don't want to give WP:INTEXT attribution to a specific author (e.g., if Koehler is not the only person to make this point, or because you are trying to comply with Wikipedia:Manual of Style/Medicine-related articles#Cite sources, don't describe them), could you fully comply with their citation style guide? WhatamIdoing (talk) 06:32, 29 July 2023 (UTC)

WhatamIdoing, I guess your examples are supposed to be from the project page. But that guideline does not contain the string "balanced news coverage can distort the public’s perception" and there is no such section "In-Text Citations". Jc3s5h (talk) 16:23, 29 July 2023 (UTC)

No, my examples are copied straight out of the APA Style Guide (current edition). The APA Style Guide says that if you are going to cite Koehler's work in support of a particular sentence, then you either need to add "(Koehler, 2016)" at the end of the sentence, or mention the author's name in the text of the article you're writing.

This style is "deprecated" (by which we mean something closer to "not allowed, but it'll take us a while to clean up existing uses") on this wiki. It is no longer "legal" to use APA Style in an English Wikipedia article. It cannot be done.

And so the question at the top become: Should WP:CITE mention, as examples of citation styles, any styles whose use is not allowed here? WhatamIdoing (talk) 17:05, 29 July 2023 (UTC)

I would describe APA style as using mostly parenthetical references, with occasional mention of the author in the flow of the sentence as an alternative. Importantly, APA style excludes using footnotes or endnotes as the method of providing inline citations. Wikipedia now calls for using endnotes for inline citations. I don't think that excludes discussing who wrote what, and when, in the flow of the text when that is relevant. Jc3s5h (talk) 17:53, 29 July 2023 (UTC)

Discussion of who wrote what, in article text, is a content issue. Citation style, or MOS in general, should have nothing at all to do with decisions about what article content is appropriate to include or not include. —David Eppstein (talk) 18:07, 29 July 2023 (UTC)

It's noteworthy that deprecation of something like "(Miller 2005, p. 1)" in a sentence is not deprecation of something like "According to Miller (2005)" when we're citing more than one Miller source.  — SMcCandlish ☏ ¢ 😼  18:44, 29 July 2023 (UTC)

Nor is it deprecation of text like "In her 2005 publication on the subject, Miller wrote..." Because both of those things are part of the text of the article, not an extratextual marker linking the text to a reference. —David Eppstein (talk) 19:52, 29 July 2023 (UTC)

And when we deprecate something like "(Miller 2005)", we effectively ban APA Style. So – should a style that isn't acceptable be mentioned in this guideline? Is there a benefit? Will it be confusing? WhatamIdoing (talk) 01:58, 30 July 2023 (UTC)

How do you add a quote from a source into a citation using a closed WP:REFNAME? For instance, adding the quote "Test" into the citation <ref name=NYT/>, so that it displays the quote just for that use of the ref name, instead of adding the quote parameter to the full reference which will display the quote in every use of that reference (so, not this: <ref name=NYT>{{cite web | ... | quote=Test}}</ref>). I have not found a workable solution across several ref templates, such as {{r}}. Lapadite (talk) 09:10, 18 August 2023 (UTC)

You can't. That's just not how it works. A <ref name="NYT" /> is just an alias/pointer to an actual citation. If you need to cite two different quotes at the same page, then you have to have two separate citations that both have the same |page= but different values for |quote=.  — SMcCandlish ☏ ¢ 😼  11:20, 18 August 2023 (UTC)

I wish there was a simple way without having to repeat full citations. Even just the ability to add text in some way to a ref name, sans a parameter, would do. At least for the common citation templates, I supposed we have to make do with repeating full citations or adding a note, which is also repeating the full citation. Lapadite (talk) 12:02, 18 August 2023 (UTC)

Use a list of the long-form sources then link the citations back to them using {{sfn}} or {{harvnb}}. Eg like at Marian reforms. Ifly6 (talk) 15:35, 18 August 2023 (UTC)

If you have a bibliography then {{sfn}} and its |loc= parameter could do it.  Stepho  talk  11:37, 18 August 2023 (UTC)

Haven't worked with sfn in a significant way, but I'll make note of that, thank you. Lapadite (talk) 12:07, 18 August 2023 (UTC)

@Lapadite: A more flexible technique than {{sfn}} is making use of the |ref= parameter of the citation template on the first citation to the source, and making a shortened second cite to the same source manually, e.g. <ref>[[#RefName|Garcia (2023)]], p. 232: "[Quotation here.]"</ref> I did a write-up for someone else about this just a little while back: User talk:SMcCandlish/Archive 199#Page-ception. I chose this over {{sfn}} in some articles I'm working on with complex citations, because {{sfn}} has few parameters, and not a quotation one; it's a "blunt instrument".  — SMcCandlish ☏ ¢ 😼  16:26, 18 August 2023 (UTC)

@SMcCandlish: This is perfect, thank you. I didn't realize the ref parameter in a citation template could be used as a customizable anchor. I'll read over your talk page post. Thanks everyone else for the suggestions. Lapadite (talk) 20:47, 20 August 2023 (UTC)

Opening a discussion of WP:REPEATCITE in response to this edit by Nightscream, which was made in response to a reversion based on REPEATCITE of Special:Diff/1171536910 at LP record. In this instance, a sentence very similar to a cited statement in earlier in the article was being used as a setup to several subsections. It was also well supported with citations in the subsections that followed. The long-standing text at WP:REPEATCITE says Material that is repeated multiple times in an article does not require an inline citation for every mention. While I understand that some editors want a reference number at the end of every single statement or paragraph, this seems a clear case where it is both unnecessary and against long-standing guidance. ——Carter (Tcr25) (talk) 18:56, 21 August 2023 (UTC)

I think that citing the same fact once per article is generally enough. The only thing I'd change about that section is removing the essay claiming that medical content needs extra copies of citations. Once you've established in an article that smoking increases the risk of lung cancer, or that HIV causes AIDS, or that chemotherapy is a treatment for cancer, you should not have to re-cite that information every time the point comes up again. WhatamIdoing (talk) 19:11, 21 August 2023 (UTC)

I don't know whether the Wikipedia medical community actually prefers the style claimed in this document, ie that they should have citations at each sentence, but inasmuch as they do we should defer thereto. Ifly6 (talk) 19:30, 21 August 2023 (UTC)

Please find my name in https://xtools.wmcloud.org/articleinfo/en.wikipedia.org/Wikipedia_talk:WikiProject_Medicine#top-editors. WhatamIdoing (talk) 20:11, 21 August 2023 (UTC)

Okay. Then you should be able to tell me whether the conditional formula I presented applies or not. Does the consensus claimed on the page, that WikiProject Medicine wants it to be that way, truthfully reflect an existing consensus? Ifly6 (talk) 21:13, 21 August 2023 (UTC)

I'd first have to say that WikiProjects (=groups of editors that want to work together to improve Wikipedia, as opposed to the many individuals who prefer to work on their own) don't get any sort of special say in the matter. WikiProjects' preferences are the "bad example" given in Wikipedia:Consensus#Level of consensus. (See also the guideline at Wikipedia:WikiProject Council/Guide#Advice pages).

Having said that, I think the opening line (i.e., that most of the community wants a greater citation density for Cancer than for Video game industry) is true. It is probably true that WikiProject Medicine (and also the rest of the community) tend to advocate for specialized rules for medicine, especially if you count creating Wikipedia:Manual of Style/Medicine-related articles and Wikipedia:Identifying reliable sources (medicine) to be "specialized rules", although the group has not advocated for this particular rule (and if we were going to, we'd recommend putting it in MEDRS or MEDMOS, not in an essay).

On the other side, the approach encouraged in Wikipedia:Citations in medical articles#Citations in the lead has been significantly and repeatedly disputed at WT:MED. It might be more accurate to say that a couple of editors not only strongly supported this but also were willing to do the work to make the articles conform to their preferences. There is no campaign at WT:MED to ban citations from the lead, but there is strong opposition to adding citations to certain/specific articles' leads, particularly FAs that were promoted without citations in the lead.

Compared to the rest of the group, on the question of "one citation per sentence", I'd say that my own views are a little less cite-y than median. (I'm also currently in second place for this year's citation-adding contest, so perhaps my practice is to add more citations than my theory requires.) WhatamIdoing (talk) 22:04, 21 August 2023 (UTC)

I think it actually serves readers better to require a citation somewhere close. The paragraph level seems reasonable. What will happen in a long article where a challengable fact is cited somewhere in one section is that the second mention (perhaps thousands of characters later) will be tagged {{citation needed}} anyway. It would be better to repeat the citation. Repeated citations are not burdensome if people use modern tool-assisted citations styles like {{sfn}}. I cannot say I appreciate editing the guidance documents without discussion to try to "win" a content dispute; it was right to revert. Ifly6 (talk) 19:30, 21 August 2023 (UTC)

In general, I'd agree that repetition of a citation is helpful in many instances, but in many cases citing basic facts once should suffice. It really depends upon what's being said and the structure (including length) of the article. ——Carter (Tcr25) (talk) 20:39, 21 August 2023 (UTC)

I agree with Tcr25 that it depends on the nature of the claim being made. Would you tag a sentence like "Smoking increases the risk of lung cancer"? I wouldn't, and I doubt that anyone on this page would. People old enough and educated enough to edit Wikipedia already know that this sentence is Wikipedia:Glossary#verifiable even if it's not Wikipedia:Glossary#cited anywhere in the article, and if they'd read the whole article and noticed a citation for that claim earlier in the article, they certainly wouldn't demand one for a second instance.

The bigger problem with "everything must be cited in every instance" is that "every instance" is not just a paragraph's worth of information, or even a sentence's. Generally, we're looking for citations to support the main idea. But sometimes, the sentence needs to contain more facts than just the main idea. Behaviors we don't want include:

• Having a well-cited section about smoking causing cancer, followed by a well-cited section on other causes that says "In addition to smoking, lung cancer may be caused by..." – and someone fact-tagging the "In addition to smoking" phrase, because the cited sources are about non-smoking causes of lung cancer, and the smoking-related citations are in the previous section (or paragraph) of the article.
• Simple summary paragraphs getting fact-tagged ("There are many causes of lung cancer", especially if such a paragraph was created to provide a simple summary of a technical section as recommended in WP:UPFRONT) because they aren't cited right there, even though the rest of the section (e.g., a l-o-n-g list of causes of lung cancer) proves the verifiability of the summary.

On the other side, we do want direct quotations are contentious BLP matter to get cited every time. If you're writing that a politician hates voters, you need to have all the little blue clicky numbers everywhere.

Editors who make a habit of tagging and blanking uncited information, including Nightscream (example), would probably find their self-chosen work less controversial and easier if citations were spammed throughout articles, so they could just skim over the article and assume that anything without a little blue clicky number was uncited and fair game for blanking. I'm not sure that anyone else really benefits from it, though. WhatamIdoing (talk) 20:48, 21 August 2023 (UTC)

@Ifly6, you have based your argument partly on the idea that readers read the sources. doi:10.1145/3366423.3380300 says this is basically not the case, especially in articles that are long enough to have disputes over whether a fact has been repeatedly cited enough times. WhatamIdoing (talk) 20:52, 21 August 2023 (UTC)

My biggest issue with the tagging and blanking behavior is that unless the editor is actually reading and editing the article along with the blanking, they end up leaving things more disjointed and less informative. And that's not even considering that a "little blue clicky number" is no guarantee of factual accuracy. Many articles at Wikipedia:List of hoaxes on Wikipedia had LBCNs, but they weren't appropriate or accurate. —Carter (Tcr25) (talk) 17:29, 22 August 2023 (UTC)

Hi! This recently came up at several FA reviews (and previously at countless GA reviews). I'm looking for some sort of consensus on how we should layout references when they are next to each other. There has been numerous requests for references to be placed in numerical order when next to each other, however, when requesting a script to do this at WP:SCRIPTREQ#REFORDER, it was brought to my attention that some users ask for the references to be in a relevance order (full details at link provided). I can't say I mind either way, but I've never put references in an order by relevance before. What I'm looking for is a consensus that we should ask for references to be in a numerical order or not to change this. Lee Vilenski ^{(talk • contribs)} 12:51, 27 August 2023 (UTC)

How about we not worry about this enough to put it into a rule? The order of the little numbers is something that we should not worry about. Ealdgyth (talk) 12:57, 27 August 2023 (UTC)

Well, the thing is that people do worry about this, well, enough to request it to be changed. If we had a consensus that it shouldn't matter what the numbers were, that would be fine. Lee Vilenski ^{(talk • contribs)} 14:15, 27 August 2023 (UTC)

Numerical order is best, but others disagree, so you won't have consensus on the issue. Headbomb {t · c · p · b} 14:16, 27 August 2023 (UTC)

Seems like the FA reviewers have too much time on their hands. Even if the references are ordered at some moment, they'll move around as the article changes. The changes are more likely in articles where certain sources are cited repeatedly at various parts of the article.

There are other criteria that might be used to order sources, such as English first, online first, or sources that anybody interested it the topic ought to own first. Jc3s5h (talk) 14:42, 27 August 2023 (UTC)

My own preference is option C: Combine all non-named citations to the same claim (including shortened footnotes) into a single reference to avoid as much as possible any refbombs of any length. {{Multiref}} is custom built for this, but if I'm cleaning up an article rather than composing, I'll usually just swap out all the </ref><ref> with {{pb}} and call it good. {{sfnp}} I'll convert to {{harvnb}}, place within <ref> tags, and separate with semicolons. Folly Mox (talk) 17:21, 27 August 2023 (UTC)

That's only practical in a very stable article. When something's still under a lot of development, doing WP:BUNDLING operations can be a serious hindrance, especially to someone else doing the actual work at the article; drive-by bundling is generally a bad idea unless you're certain that most of the work on the article is already done.  — SMcCandlish ☏ ¢ 😼  18:18, 27 August 2023 (UTC)

I agree, and almost none of my cleanup is on active articles (in fact, it's usually only necessary because the article has been neglected by interested contributors). If I have a counterpoint, it's that rearranging adjacent citations based on numerical order is even less suitable for articles in active development. Folly Mox (talk) 19:36, 27 August 2023 (UTC)

Yep. Stuff is apt to move around a lot.  — SMcCandlish ☏ ¢ 😼  20:27, 27 August 2023 (UTC)

Lee Vilenski is correct that the guideline needs to address this one way or another (probably by stating that there is no consensus on a specific ordering, and that editors should not edit-war over it or make willy-nilly changes to ordering to suit their personal preferences). This would resolve the problem the editor has brought to us, of frequent FAC/GAN demands (based on no actual guidance or other consensus) for such ordering changes. And it would forestall WP:MEATBOT-style futzing around with ref order, which is something many of us have encountered as a pointless watchlist trigger.  — SMcCandlish ☏ ¢ 😼  18:18, 27 August 2023 (UTC)

Agreed, reflist order is not something I much care about and we can usefully forestall arguments by clearly stating as much.--Sturmvogel 66 (talk) 19:15, 27 August 2023 (UTC)

I agree -- it would be best if this guideline made it clear there is no requirement of this kind. Mike Christie (talk - contribs - library) 19:29, 27 August 2023 (UTC)

Doesn't it already? See TSI: "References need not be moved solely to maintain the chronological order of footnotes as they appear in the article". Nikkimaria (talk) 19:41, 27 August 2023 (UTC)

That'll teach me to make assumptions. Lee, I think that wording does indeed settle it -- were you unaware of it, as I was? Mike Christie (talk - contribs - library) 20:32, 27 August 2023 (UTC)

I wasn't aware of that wording, perhaps we should re-redirect WP:REFORDER to that part. Lee Vilenski ^{(talk • contribs)} 20:37, 27 August 2023 (UTC)

Well I'll be damned! It was buried in the middle of unrelated text about not changing the position of the citations in relation to the material they support. I've fixed that and added a WP:CITEORDER shortcut [1], and this is probably enough to resolve Lee Vilenski's reported issue. (I was actually doing this while Lee was posting that!)  — SMcCandlish ☏ ¢ 😼  20:38, 27 August 2023 (UTC)

PS: Yes, it would probably be good to redirect WP:REFORDER there as well, though that's probably something to propose at WP:RFD rather than just usurp it boldly.  — SMcCandlish ☏ ¢ 😼  20:44, 27 August 2023 (UTC)

Opened the RfD.  — SMcCandlish ☏ ¢ 😼  20:52, 27 August 2023 (UTC)

Shouldn't that read "numerical" rather than "chronological"? -- Valjean (talk) (PING me) 20:47, 27 August 2023 (UTC)

Yes, good point. Will tweak it in a sec.  — SMcCandlish ☏ ¢ 😼  20:52, 27 August 2023 (UTC)

Based on the positive feedback above, I also made the wording more general. If that's too much too soon, feel free to revert.  — SMcCandlish ☏ ¢ 😼  21:00, 27 August 2023 (UTC)

Thanks for all this. This seems like plenty to have a discussion/policy to point to when suggested. Lee Vilenski ^{(talk • contribs)} 16:08, 28 August 2023 (UTC)

Some pages have come up with very creative solutions for this. Check out Bubsy 3D#Notes for example. As Folly Mox says, many pages bundle citations. I don't see a policy on reforder being helpful, Rjjiii (talk) 19:42, 27 August 2023 (UTC)

Sure, that's more of a WP:BUNDLING to solve WP:CITEKILL though. I don't think we should be bundling two citations together for the hell of it. Lee Vilenski ^{(talk • contribs)} 20:02, 27 August 2023 (UTC)

Only if refs are for the exact same piece of info. Otherwise, accurate and exact placement (word, phrase, or sentence) is far more important than numerical placement. -- Valjean (talk) (PING me) 20:45, 27 August 2023 (UTC)

'Twould probably be better done in that case using the template recommended at WP:BUNDLING, instead of arguably misusing the {{efn}} template, which is for informational footnotes, not citations.  — SMcCandlish ☏ ¢ 😼  21:10, 27 August 2023 (UTC)

As an exposition – which nobody asked for – it's not my MO to zip around article to article and bundle every pair of adjacent citations I find. And I don't take citation bundling as a primary cleanup impetus. (Do you like my strawmans?) It's usually some other form of citation cleanup, typically in the wake of Citoid and affiliates. If I'm already doing substantive cleanup, and an article has a problem with long strings of adjacent citations, I'll bundle them as I go in a kind of GENFIX fashion.

And I do take Valjean's point just above about precise placement taking priority. Unless it's an article in a topic I'm legitimately interested in and I want to leave it pristine like a kitchen after closing, I'm not in the actual sources verifying claims and assessing due weight and everything. I'll increment the referencing quality without going full professional, and move on to the next of the thousands and thousands of problem articles in whichever cleanup queue I'm working. Folly Mox (talk) 21:00, 27 August 2023 (UTC)

Yes, they are fine mans full of fresh straw. :-) That sounds like a sensible approach. For my part, at Regimental tartan and related articles, there's a thick stack of cites on the sentence addressing the overwhelmingly common myth that clan tartans are "ancient" (they really date almost entirely from 1815 onward). I've not bundled them yet because the material's still under construction (I've only used about half the sources I have yet); some cites to add there may be better as replacements rather than additions; some of the cites are to sources used multiple times in the same piece, so need to be reconstructed properly before bundling; the similar material in the related articles is not all citing the same sources (e.g. the regimental one is citing mostly the ones that specifically relate regimental tartans ancestrally to clan tartans, while the main tartan article and some forthcoming split-offs of that over-long article will cite more general sources on the topic, including ones that also discuss district tartans as ancestral to clan ones); and so forth. It's just not stable yet, but I'm keenly aware that they need eventually to be bundled. If someone just went and bundled them right now I would revert them, as impeding the work I'm doing (and doing alone – no one's substantively touched the topic in years, and the related wikiprojects for Scotland and clans are moribund).  — SMcCandlish ☏ ¢ 😼  21:10, 27 August 2023 (UTC)

If someone came along and bundled cites on an article I was the main editor of, I'd probably either revert or post a note to the editor asking if there was a specific reason. I'm not keen on bundled cites; I wouldn't unbundle someone else's cites but I think it's a choice that the editors working on the article should be allowed to make. Mike Christie (talk - contribs - library) 21:16, 27 August 2023 (UTC)

In the type of cleanup I typically find myself doing, no clear thought seems to have gone into the citations. Sometimes it's a single claim that someone felt the need to cite twelve or fifteen times in response to some POV pusher years past, or a result of POV pushers themselves, dropping in a half dozen bare URL direct page gbooks links which some other editor later drives by using a script to expand into inaccurate citation templates. It's a mess out there.

The only time I found myself really tinkering with citations that were intentional was at an article that had a weird bifurcation between alphabetic and numeric footnotes in a way that left no space for explanatory footnotes, had alphabetic footnotes running to "gt" and one source cited in so many {{sfnp}}s that the letters to hop back to the place it was cited from the reference section went all the way to "bl", which I think means 90 in thirtysixidecimal.

Which I think is a long way of trying to reassure people here that I don't find myself changing the intent of primary contributors so much as doing some organisation where previously there was none discernible. Folly Mox (talk) 22:00, 27 August 2023 (UTC)

Thanks -- that's a helpful explanation. Mike Christie (talk - contribs - library) 22:02, 27 August 2023 (UTC)

I want to thank the three editors offering concerns / feedback / pushback on my gung-ho bundling style here. Even though I doubt articles where the main authors have been active on the article within the past year or so will wind up in the cleanup queues I work, and citation bundling is a relatively infrequent consequence of my gnomework, I'll double check for recent activity before I put my bundling pants on. Thanks for the reminder of conscientousness. Folly Mox (talk) 23:29, 27 August 2023 (UTC)

A recent discussion at a FAC led me to realize that there's a difference in phrasing amongst WP:V, WP:CITE, WP:NONFREE, and WP:CLOP regarding in-text attribution. V and NONFREE are policy, CITE is a guideline, and CLOP is an essay. I originally posted a note at the village pump about changing CLOP, but I now think it would make more sense to resolve the conflict between the two policies and the guideline first. With those three in sync it should be easy to agree on a harmonizing change to CLOP.

Here's the relevant wording:

• V says Summarize source material in your own words as much as possible; when quoting or closely paraphrasing a source, use an inline citation, and in-text attribution where appropriate.
• NONFREE says use brief verbatim textual excerpts from copyrighted media, properly attributed or cited to its original source or author (as described by the citation guideline).
• CITE says In-text attribution should be used with direct speech (a source's words between quotation marks or as a block quotation); indirect speech (a source's words modified without quotation marks); and close paraphrasing. It can also be used when loosely summarizing a source's position in your own words, and it should always be used for biased statements of opinion.

The two policies do not define when to use in-text attribution instead of just citation; I take this (and the use of "where appropriate") to mean that it's up to editorial discretion. The distinction between should be used and should always be used in CITE doesn't make sense to me unless it's interpreted similarly: that is, the former allows for some editorial discretion whereas the latter does not. I would like to change the text quoted above to In-text attribution may need to be used with direct speech (a source's words between quotation marks or as a block quotation); indirect speech (a source's words modified without quotation marks); and close paraphrasing. It may also be used when loosely summarizing a source's position in your own words, and it should always be used for biased statements of opinion. Mike Christie (talk - contribs - library) 18:31, 27 August 2023 (UTC)

• Support. This is clearly a sensible resolution to a minor but confusing WP:POLICYFORK problem, and it would better match actual practice.  — SMcCandlish ☏ ¢ 😼  18:50, 27 August 2023 (UTC)
• Support. Unfortunately many of the policy documents have unclear usages of 'should'. I will say 'must' is clearer than 'should always', and means the same thing. So I would also support keeping the first should and changing the second to 'must' (from 'should always'). -- LCU ActivelyDisinterested ∆transmissions∆ °co-ords° 19:03, 27 August 2023 (UTC)
• Support per the rationale of the three above editors. - SchroCat (talk) 19:12, 27 August 2023 (UTC)
• Support - a reasonable change. I also agree with ActivelyDisinterested that replacing "should always" with "must" would be an improvement. — Golden ^talk 19:21, 27 August 2023 (UTC)
• Support I don't consider there to be any distinction between "should" and "should always" and regard "always" as an unnecessary noise word. I would strongly urge that the MOS standardise the use of these words along the lines of RFC 2119 to avoid confusion in the future:

  Must means that something is absolutely required by policy

  Should means that something is recommended. There may exist valid reasons in particular circumstances to ignore it, but the full implications must be understood and carefully weighed before choosing a different course

  May means that something is truly optional.

  Hawkeye7 (discuss) 19:37, 27 August 2023 (UTC)

  Just FYI, WP adopting RFC 2119 wording and definitions has been proposed before and failed to gain consensus. That said, we can certainly massage the text in that general direction just for clarity's sake, without being "official" about it.  — SMcCandlish ☏ ¢ 😼  21:14, 27 August 2023 (UTC)

  This was my point of when I said that "must" should be used in this case instead of "should always". "Should always" is commonly used to mean "must", but that's not the strict sense of the word. -- LCU ActivelyDisinterested ∆transmissions∆ °co-ords° 21:49, 27 August 2023 (UTC)

  Well, this is a guideline, not a policy, so it should never say "must" about anything unless it's citing a policy that requires it.  — SMcCandlish ☏ ¢ 😼  16:14, 29 August 2023 (UTC)

• Support because it's confusing to have slightly different policies on different pages, Rjjiii (talk) 19:44, 27 August 2023 (UTC)

I've gone ahead with this wording change. I used "should always" rather than "must". I see there's some support for the latter, but I feel it's better to use the text everyone commented on. I very slightly prefer "should always" myself but if someone were to change it to "must" I wouldn't object. Mike Christie (talk - contribs - library) 11:18, 29 August 2023 (UTC)

I have now suggested a corresponding change to the close paraphrasing essay, here. Mike Christie (talk - contribs - library) 11:47, 29 August 2023 (UTC)

• Oppose and regret that you've gone ahead after such short discussion. The WP:INTEXT original "should be used" is now "may need to be used", which is another way of saying "la la la we no longer care". The original wording did not contradict WP:V, the WP:V words "where appropriate" were clarified by saying here's a guideline that it's appropriate. Plus, there is a reference from WP:ATTRIBUTEPOV: "Biased statements of opinion can be presented only with in-text attribution." -- because of that, biased statements "should" require in-text attribution, but suddenly that's not so any more. Plus, MOS:QUOTEPOV starts with "Quotation should be used, with attribution, ..." notice the word "should" which used to correspond with the word "should" in WP:INTEXT, now due to your change it doesn't correspond. Plus, the WP:QUOTE essay's words "POV language must be quoted and attributed" are now obsolete because an essay that doesn't match a PAG is junk. Plus, there might be a slight weakening of the WP:RS/QUOTE requirement "the text of quoted material is best taken from (and cited to) the original source being quoted" because when the prior WP:INTEXT wording is followed that meant you were not merely supposed to cite you were supposed to attribute. Peter Gulutzan (talk) 15:24, 29 August 2023 (UTC)

  It was a short discussion, but not a formal RfC, and without opposition I didn't think there was a need to wait. Let's see if there are other similar comments. Mike Christie (talk - contribs - library) 15:32, 29 August 2023 (UTC)

  "may need to be used" clearly cannot mean "we don't care", because "need" indicates a requirement. We're just indicating a conditional requirement, i.e. one that depends on the context (that is, some contexts will require it, some will not). There may be some other way to phrase it, but the point of all this was to stop implying in one place that it was up to editorial discretion ("where appropriate") and implying in another that it was not.  — SMcCandlish ☏ ¢ 😼  16:16, 29 August 2023 (UTC)

  "may need to be used" clearly means "we don't care" because "may" indicates a non-requirement, which is also what your wording "editorial discretion" means, which is not what "where appropriate" meant till now. Plus, WP:PLAGIARISM has yet another reference to what the gutted guideline used to contain: "INTEXT: Add in-text attribution when you copy or closely paraphrase another author's words or flow of thought, unless the material lacks creativity or originates from a free source." Plus, now WP:INTEXT doesn't correspond to Mos:Attribution "The source must be named in article text if the quotation is an opinion ..." (not "may" and not merely "biased opinion"). Incidentally "where appropriate" was added by Kotniski without a discussion that I could see, although shortly later "Proposed wording re in-text attribution" was discussed in an old WP:V talk page. Peter Gulutzan (talk) 16:22, 31 August 2023 (UTC)

  MoS (and other guidelines) use "may" and other conditional wording all over the place. They exist not as a list of hard-and-fast rules, like our legal policies, but often serve the purpose of forestalling editwars by indicating what is permissible, what is subject to editorial discretion, i.e. what should be consensus-discussed on an article-by-article basis. That doesn't mean no one cares, it means we care more about stopping disruption on that particular matter than about imposing an abitrary bright-line rule. If we have some new conflict between guideline wording, then of course we should fix it.  — SMcCandlish ☏ ¢ 😼  21:05, 31 August 2023 (UTC)

  I'm not sure I like may need to without some sense of when or why that need might arise: it's not quite the same as saying x may be used, which is equivalent to "it's up to you". If we're saying that a need may arise, we should have some way of demonstrating that such a need has arisen. Suggest that the litmus test here is some combination of how close the paraphrasing is and how far the paraphrased material moves beyond bare facts. There's also a matter of how debatable the quoted material is and how it's couched in the sentence as a whole: Augustus has been called a "terrible emperor" is fine, but taking the attribution out of According to Smith, Augustus was "a terrible emperor" would not. However, setting all this out in too much detail would risk major instruction creep... UndercoverClassicist ^T·C 10:04, 1 September 2023 (UTC)

WP:WHENNOTCITE says not to repeat the same citation on consecutive sentences in the same paragraph. WP:DYK requires that the same citation be repeated, when needed to source a hook claim that appears in the first of the two sentences. Delaying the same citation until the second of two consecutive sentences is not allowed. Is this contradiction a problem? —David Eppstein (talk) 06:20, 30 August 2023 (UTC)

I don't read it as a problem. WP:WHENNOTCITE uses the 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. It is not necessary to include a citation for each individual consecutive sentence, which sounds like it leaves space for exceptions and carve-outs. Folly Mox (talk) 06:30, 30 August 2023 (UTC)

Well, any guideline has exceptions (think WP:IAR). However, what makes DYK so unique that a reader cannot figure out that a later citation supports earlier sentences?—Bagumba (talk) 07:17, 30 August 2023 (UTC)

@Bagumba, people at DYK who check multiple hooks in a short period (promoters and movers-to-queue) need that sentence cited, particularly if the full passage ends with multiple citations, in order to check the source efficiently. Even with the hook sentences directly cited, it can take significant time to do the necessary rechecks of 8 hooks. If they have to go looking in multiple sources, the time required mushrooms. Valereee (talk) 12:33, 9 September 2023 (UTC)

Whoops, sorry, didn't realize I was replying to a discussion that was 9 days old! Valereee (talk) 12:34, 9 September 2023 (UTC)

@Valereee: I think the OP was referring to a different scenario, where the hook is in sentence X, but the source for sentence X also supports sentences X+1, X+2...Y. Typically, a non-DYK page would put the single citation after sentence Y. In your scenario, where multiple sources are cited at sentence Y, it seems that ideally the DYK nom's "source" parameter would provide details on the specific source and provide a quoted excerpt to isolate which source(s) among the multiple citations actually support the hook. —Bagumba (talk) 14:37, 9 September 2023 (UTC)

Yeah, I definitely get that, it's just that the rules at DYK are complex enough that explaining that it's okay to have the citation for the hook sentence be at the end of the paragraph if the paragraph is entirely sourced to that and only that source, the hook sentence doesn't need a citation, but if the para has several citations and the hook sentence is sourced only to one of them (or to two of the four, or whatever) then the hook sentence needs a direct citation...well, it's just easier for everyone to understand that the hook sentence needs a direct citation. Experienced reviewers/promoters/movers can IAR that in those cases where the direct cit isn't strictly necessary. :D And in the case of experienced nominators, it's also just easy enough to wait until the silly thing has appeared, then go back and remove that citation if it's going to bother you. Valereee (talk) 14:46, 9 September 2023 (UTC)

Interesting background if that's the origin of the rule. I've done a fair share of DYKs, and previously thought it was obtuse to repeat a citation because of that DYK rule. But I had only seen it come up when there was only a lone citation, and it was obvious that it was a few sentences later. And it seemed bureaucratic to add the source, purely for DYK, just to remove it later (if someone was bugged by it) and still be guideline compliant. —Bagumba (talk) 15:02, 9 September 2023 (UTC)

I don't actually know the origin, but that's certainly the reason promoters and movers make the strongest arguments for whenever anyone complains that it's nitpicky. :D It is absolutely infuriating to read an entire source because it's the first one to follow the hook sentence (maybe at the end of the following sentence), search it multiple times for a combination of wordings, then ping the nom and have them say, "Oh, that's actually from source 3 at the end of the para." Valereee (talk) 15:34, 9 September 2023 (UTC)

Honestly, the DYK carve-out has existed for so long that there's a pretty clear community consensus for it. This conflict only exists on paper, the community is fine with DYK requiring consecutive citations. It's perfectly within consensus to just add a footnote here re: DYK. theleekycauldron (talk • she/her) 07:58, 30 August 2023 (UTC)

That would be wise, so this doesn't ever come up again, and because we really don't tolerate WP:POLICYFORKs. If two sets of rules here seem to contradict each other, the solution is to erase the conflict one way or another, not let it continue.  — SMcCandlish ☏ ¢ 😼  08:18, 30 August 2023 (UTC)

Adding a footnote here for a DYK exception seems like a good simple solution to me. —David Eppstein (talk) 19:15, 30 August 2023 (UTC)

I've added a footnote, anyone should feel free to correct whatever I did wrong! Valereee (talk) 15:36, 9 September 2023 (UTC)

I think the addition is unnecessary, because the premise is false. WP:WHENNOTCITE doesn't actually say not to repeat the same citation on consecutive sentences in the same paragraph, as alleged above. It says that repeated citations are not necessary, but it does not say they're not permitted. WhatamIdoing (talk) 02:44, 13 September 2023 (UTC)

I do not think a DYK exception is needed. Instead, DYK should simply refer to existing guideline WP:INTEGRITY:

...adding text without clearly placing its source may lead to allegations of original research, of violations of the sourcing policy, and even of plagiarism.

—Bagumba (talk) 04:31, 13 September 2023 (UTC)

Referring to the usual practices does not address DYK's unusual need, which is for the organizers to be absolutely certain that there is an inline citation for whichever fact is appearing on the Main Page. WhatamIdoing (talk) 01:38, 14 September 2023 (UTC)

...for the organizers to be absolutely certain that there is an inline citation for whichever fact is appearing on the Main Page The only way to be certain, regardless of the position of the citation, is to check the source. It could just as well be that the hook is supported by sentence N, but sentence N's citation supports sentence N-1 and sentence N-2, but still not sentence N.—Bagumba (talk) 04:34, 14 September 2023 (UTC)

Sorry, I'm not following...yes, the source needs to be checked, and having the relevant cit on the hook sentence makes it easier/more efficient to check to make sure the source does indeed support both the hook and the relevant sentence/s in the article. Valereee (talk) 14:13, 14 September 2023 (UTC)

Example Seems like we are discussing the following scenarios:

1. Sentence with hook.^[1] Sentence after sentence with hook.^[1]
2. Sentence with hook. Sentence after sentence with hook.^[1]

Is case 2 that confusing? The onus is on the nominator and approver to ensure source WP:INTEGRITY is met. This is standard verification procedure outside of DYK.—Bagumba (talk) 15:39, 14 September 2023 (UTC)