Open main menu

Wikipedia β

Wikipedia talk:Citing sources

Wikipedia Help Project (Rated NA-class, High-importance)
WikiProject icon This page is within the scope of the Wikipedia Help Project, a collaborative effort to improve Wikipedia's help documentation for readers and contributors. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks. To browse help related resources see the Help Menu or Help Directory. Or ask for help on your talk page and a volunteer will visit you there.
 NA  This page does not require a rating on the project's quality scale.
 High  This page has been rated as High-importance on the project's importance scale.
 
Frequently asked questions (FAQ)
Why doesn't Wikipedia require everyone to use exactly the same style for formatting citations on every single article, regardless of the subject?
Different academic disciplines use different styles because they have different needs and interests. Variations include differences in the choice of information to include, the order in which the information is presented, the punctuation, and the name of the section headings under which the information is presented. There is no house style on Wikipedia, and the community does not want to have the holy war that will happen if we tell people that they must use the style preferred by scientists in articles about history or the style preferred by artists when writing about science. Editors should choose a style that they believe is appropriate for the individual article in question and should never edit-war over the style of citations.
What styles are commonly used?
There are many published style manuals. For British English the Oxford Style Manual is the authoritative source. For American English the Chicago Manual of Style is commonly used by historians and in the fine arts. Other US style guides include APA style which is used by sociologists and psychologists, and The MLA Style Manual which is used in humanities. The Council of Science Editors and Vancouver styles are popular with scientists. Editors on Wikipedia may use any style they like, including styles they have made up themselves. It is unusual for Wikipedia articles to strictly adhere to a formally published academic style.
Isn't everyone required to use clickable footnotes like this[1] in every single article?
Footnotes (also called "<ref> tags") are popular but not required. The purpose of an inline citation is to provide information about where that material came from. Any system that allows someone to figure out which source supports which material achieves that goal and is therefore acceptable. Other styles, such as parenthetical citations, are simpler for new users to understand, are commonly taught in schools, and may be the style preferred by the relevant academic discipline.
Why doesn't Wikipedia require everyone to use citation templates in every single article?
Citation templates have advantages and disadvantages. They provide machine-readable meta data and can be used by editors who don't know how to properly order and format a citation. However, they are intimidating and confusing to most new users, and, if more than a few dozen are used, they make the pages noticeably slower to load. Editors should use their best judgment to decide which format best suits each specific article.
Isn't there a rule that every single sentence requires an inline citation?
No. Wikipedia:Verifiability requires citations based on the content rather than the grammar. Sometimes, one sentence will require multiple inline citations. In other instances, a whole paragraph will not require any inline citations.
Aren't general references prohibited?
A general reference is a citation listed at the end of an article, without any system for linking it to a particular bit of material. In an article that contains more than a couple of sentences, it is more difficult to maintain text-source integrity without using inline citations, but general references can be useful and are not banned. However, they are not adequate if the material is one of four types of content requiring an inline citation. The article Early life of Joseph Smith, Jr. is an example of a featured article that uses some general references.
Can I cite a sign?
Yes, signs, including gravestones, that are displayed in public are considered publications. If the article is using citation templates, then use {{cite sign}}. You may also cite works of art, videos, music album liner notes, sheet music, interviews, recorded speeches, podcasts, television episodes, maps, public mailing lists, ship registers, and a wide variety of other things that are published and accessible to the public.
WikiProject Manual of Style
WikiProject icon This page falls within the scope of WikiProject Manual of Style, a drive to identify and address contradictions and redundancies, improve language, and coordinate the pages that form the MoS guidelines.
 

Value of ISSNsEdit

Editors following this page may be interested in WP:VPPOL#Including ISSNs in citations. Consider commenting there. --Izno (talk) 14:51, 10 June 2018 (UTC)

Bundling citationsEdit

May be it worth to leave a link to the relevant (IMHO) Wikipedia:Nesting footnotes article in this section? DAVRONOVA.A. 13:44, 11 June 2018 (UTC)

Or not? The very concept of "nesting footnotes" is dubious, and I think we should not be encouraging it. ♦ J. Johnson (JJ) (talk) 20:37, 11 June 2018 (UTC)

Citing a web page with no URL?Edit

I want to cite a result page from performing a search on https://kodiapps.com/search.php. Unfortunately, it is a stupidly designed site and there doesn't appear to be any URL that gets you directly to the search result. How should I cite it? -- RoySmith (talk) 12:46, 14 June 2018 (UTC)

I would use the same approach generally used in external citation guidelines, such as Chicago Manual of Style when citing a dictionary or encyclopedia. Rather than giving a page number, the entry to search for is given, after the abbreviation s.v., which stands for sub verbo, "under the word". So the parameter would be |at= s.v. "Perry Mason" if that is who you were looking up. Jc3s5h (talk) 13:00, 14 June 2018 (UTC)
How about not using obscure Latin abbreviations that very few readers will understand? The principle of listing the web page and the search term is good but there is no need to duplicate the paper-saving practices of specialist journals. Spell it out in English. —David Eppstein (talk) 13:24, 14 June 2018 (UTC)
I'd go simply
or
Headbomb {t · c · p · b} 13:32, 14 June 2018 (UTC)
|at=Found at the result starting "Banana Man…" in the search results for "Foo bar *.?" Fifelfoo (talk) 11:14, 17 June 2018 (UTC)
A "web" page always has a url, or you would not be able to access it. What you are really trying to cite is the search result. In many cases the search terms are encoded in the url – e.g., look at any Google Books url – so that can be recorded. Which is a good idea, as I vaguely recall a dispute years back where two people were getting different search results, even though both entered the same terms in the search box. It seems one person's browser was throwing in some extra parameters, which were found only upon comparing the entire search url.
I would be cautious about "citing" a search result. It's okay as an index of where to find a source, but as a source in itself (perhaps as authority for something like "78% of blonds prefer ..."?) it's just too soft to rely on. Even if the underlying database is reliable enough, it is conceivable that some search terms could return results not endorsed by the "authors", and so would verge on OR. ♦ J. Johnson (JJ) (talk) 19:34, 16 June 2018 (UTC)
"How about not using obscure Latin abbreviations that very few readers will understand?" is easily dealt with by providing wikilink assistance for such readers. E.g., S.v.. Wtmitchell (talk) (earlier Boracay Bill) 03:35, 18 June 2018 (UTC)
"S.v." is virtually never used in our citations. It should not be used in our templated ones, because we have the |at= parameter for this, e.g. {{Cite book|title=Collins English Dictionary|at="antidisestablishmentarianism" entry|...}}. It's also common and permissible to simply do {{Cite book|title=Collins English Dictionary|...}}; this is preferable, because it also permits the inclusion of the |page= or |pages= parameter (with which |at= is mutually exclusive).  — SMcCandlish ¢ 😼  10:30, 18 June 2018 (UTC)

Citing a govenment document - Specifically a RCAF Organization OrderEdit

I am working on the page RCAF Station Vulcan where was some misinformation on the page and I have found an original source document that clarifies the issue. http://heritage.canadiana.ca/view/oocihm.lac_reel_c12325/457?r=0&s=6 is the web address this document is published at but really it is only displayed on the website. The document itself is RCAF Organizational Order No. 204, File Number 925-247-1 (D . of O.) dated 3 June 1942. What is the best way to site this source document? Mech1949 (talk) 00:13, 15 June 2018 (UTC)

I would use {{cite}} to make the following: RCAF Organizational Order No. 204, Royal Canadian Air Force, 3 June 1942, File Number 925-247-1 (D . of O.), retrieved 16 June 2018 – via Héritage   Stepho  talk  10:02, 16 June 2018 (UTC)
Thank you for the recomendation, this looks much better than what I had put together.Mech1949 (talk) 15:43, 16 June 2018 (UTC)
Yeah, but don't use "D. of O." per MOS:ABBR. That string is meaningless to most readers.  — SMcCandlish ¢ 😼  10:31, 18 June 2018 (UTC)
SMcC, the "D. of O." string is a part of the file reference, see top right of the cited document. I'm guessing that it is the originating body (I've found a reference to Australia/NZ using D of O for "Director/Directorate of Organisation") and that file number 925-247-1 could be reused by another body. Martin of Sheffield (talk) 11:16, 18 June 2018 (UTC)
If it's not in the title or work parameters, it should be wikimarkup-capable, I think, so we can use {{abbr}} to explain it.  — SMcCandlish ¢ 😼  17:03, 18 June 2018 (UTC)
Such as: id=File Number 925-247-1 ({{abbr|D. of O.|Director/Directorate of Organisation}}) which gives
RCAF Organizational Order No. 204, Royal Canadian Air Force, 3 June 1942, File Number 925-247-1 D. of O., retrieved 16 June 2018 – via Héritage  ? That's assuming that the Canadian meaning of D of O is the same as the Australian one of course! Martin of Sheffield (talk) 17:21, 18 June 2018 (UTC)

Citing booksEdit

I am in the middle of a disagreement with an editor who believes that it's preferable in this instance to cite a whole book to cover about four or five paragraphs, and without citing page numbers. Perhaps it's my memory failing me, but I had remembered reading a guideline which required citing pages numbers or chapters with books. Now I see WP: Citing sources#What information to include states in the ninth bullet point, "chapter or page numbers cited, if appropriate." When would it not be appropriate? How is a third party to verify a reference to a book without narrowing the scope to at least a range of pages or a chapter? The second issue is how many citations? I am not aware that there is such a guideline, but I generally sense there is something wrong when there not at least one citation per paragraph.

The book in question is 223 pages. If I am right, what policies support this? Cheers, Oldsanfelipe (talk) 21:18, 18 June 2018 (UTC)

WP:BURDEN, part of WP:V has this: "Cite the source clearly and precisely (specifying page, section, or such divisions as may be appropriate)."
Trappist the monk (talk) 21:32, 18 June 2018 (UTC)
Namely, the word "precise". Citing an entire book for 5 paragraphs of information is not a precise citation. --Izno (talk) 22:11, 18 June 2018 (UTC)
That formulation "if appropriate" can be a misleading as specifying page numbers or at least a chapter is the default case for normal sourcing. But note that these format and template descriptions are also used for so called "general references" and "further reading" or selected bibliography sections and for those cases a specification down to chapters or pages usually doesn't make sense, i.e. is not appropriate.--Kmhkmh (talk) 22:24, 18 June 2018 (UTC)
Thanks everybody. I don't know the policy details as I should. On the other hand, the "if appropriate" clause seems ripe for abuse. I think I will appeal to the general idea of making it easier for other editors to check our work, plus the "clear and precise" language. Cheers, Oldsanfelipe (talk) 22:35, 18 June 2018 (UTC)
I agree that "if appropriate" can be misread or abused and don't mind it to be rephrased of others feel the same way. Maybe simply adding a footnote to the "if appropriate" phrase stating that usually for content sourcing page numbers or at least a chapter specification is required.--Kmhkmh (talk) 10:05, 19 June 2018 (UTC)
I don't have a clear idea yet of how to express this, but I think "appropriate" should be understood to mean the smallest unit (page range, section, chapter, etc.) available that provides support for the specific contents of a sentence or paragraph in the WP article. This may sometimes mean that adjacent paragraphs have citations to overlapping page ranges. We need to make it as easy as possible for readers to find the relevant portions of a cited source. As I said, I'm not yet sure how to state this in a non-convoluted way. - Donald Albury 11:57, 19 June 2018 (UTC)
I feel the need to point out that WP:BURDEN does not use the phrase if appropriate. There is no if in the sentence that I quoted. The parenthetical clause is merely shorthand to avoid listing many possible divisions (as was done at Template:Cite book#In-source locations).
Trappist the monk (talk) 12:30, 19 June 2018 (UTC)

──────────────────────────────────────────────────────────────────────────────────────────────────── (edit conflict)The problem is that using a full inline citation for each reference is a PITA. Using short references or named references to books already cited looks ugly, in 50 or 100 references where is the main one? There are a number of approaches which overcome this:

  • Use List-defined references (WP:LDR), see Royal Oak, Frindsbury for an example. Where a book is involved you will probably need to use either {{r}} or {{rp}} to give the page numbers, or else add the abbreviated references with pages after the extended ones.
  • Cite all the books in a separate list and then use {{sfn}} to link a specific page reference to the citation. See St. Mary & St. George Anglican Church for a straightforward example or Subhas Chandra Bose for a complex one involving notes, references and citations. Martin of Sheffield (talk) 12:38, 19 June 2018 (UTC)
Basically what I've been doing for a while, now: Francisco Menéndez Márquez. - Donald Albury 13:12, 19 June 2018 (UTC)
(edit conflict)See your talk page for a comment on this. Martin of Sheffield (talk) 13:53, 19 June 2018 (UTC)

I think this leaving the scope original question and mixing in highly controversial and somewhat subjective issues about citation styles. I think we should focus on simply making clear to readers, that "if appropriate" is not to be understood as it being ok to skip a more detailed specification (page numbers, chapters) when sourcing Wikipedia content with book citations. There seems to be a (long practised and established) general consensus for that. For the question however, how the specification should be done in detail, that is in which format, there is afaik no real consensus whatsoever, on the contrary it is a long standing controversial issue.--Kmhkmh (talk) 13:18, 19 June 2018 (UTC)

Yes, there are many acceptable styles of citation, and I tried several before settling on what I am comfortable with, but I think we should encourage editors to provide citations that make it as easy as possible for readers to verify that a source does support the statements in an article. In my opinion, that means using page numbers when available, and if not, the lowest level division or divisions in a source that support the text of an article. Failing to be as specific as possible in a citation is a disservice to the readers, at least whatever proportion of readers care to check the citations. - Donald Albury 13:48, 19 June 2018 (UTC)
(edit conflict)I think your first sentence is a little strong, given that I showed two different ways of doing things, and mentioned a third. What is germane is that simply repeating full citations inline leads to a mess such as this old version which is the sort of clutter that I assume HGFriedman is concerned about. Martin of Sheffield (talk) 13:53, 19 June 2018 (UTC)
Well there is nothing wrong with making suggestions how to facilitate a more specific citation. However the original question was not about that. The was no issue with how to provide a page number just with editor declining to provide one.
I wanted to avoid this thread moving into the pros and cons of your suggestions as they are part of an longtime unresolved wider issue. And as great or superior they might appear to you, they might do so to others. There is a reason this field is longtime contentious and imho people don't even really what is annoying clutter and what not and all the suggested templates have supporters and opponents. I for instance intentionally use none of those templates (and for good reasons from my perspective).--Kmhkmh (talk) 14:18, 19 June 2018 (UTC)
I have notified HGFriedman since his name has come up. @Martin of Sheffield: I never raised the issue of citation style. I asked about when and how often a citation should be required or appropriate. As matter of policy, it is uncontroversial that (aside from the lede text) every statement must have a source. So the clutter in the article you cite is caused by different statements backed up by different sources. I raised the issue of a single book citation with no page references to backup a handful of paragraphs. My intuitions told me this was very wrong, but my knowledge of WP policy details did not match my confidence in my own intuitions. So I asked for help. Though this is not a WP policy, for my own rule for editing, I provide no less than a citation per paragraph. However, when a single source supports all of the claims in the paragraph, I use only one citation. Therefor, I am not advocating for clutter either.
In addition to learning some details about citation policies from this thread, I have come to the conclusion that I should improve my own citation practices. For example, IIRC, WP:CIT says that it is not necessary to include page numbers in a citation of an article. I have used that to follow many others who do not cite specific page number or ranges for articles. However, with a long article of 25 to 50 pages, I have not been serving my readers well. I will start citing page numbers for long articles. Thanks again, Oldsanfelipe (talk) 16:49, 19 June 2018 (UTC)
There's a few people putting words into my mouth, or at least misunderstanding my points. I would agree that page numbers (or similar) should always be used. I do understand though that if references are not done carefully then they do degenerate into a clutter, which is why I assume that HGFriedman was not using page numbers and allowing all the references to coalesce (see Talk:Streetcars_in_New_Orleans#Citing_sources). I do think that endlessly repeating a full citation with only a page number change is fruitless, as shown in this example. You'll notice that the first source is repeated at length 7 time, the third 11 times! If editors come here asking questions I assume that they are seeking information, and that is why I gave three different ways of doing it (named references, List-defined references and Harvard). For the record I do prefer data normalisation and ordering, so would choose one of the latter two, but have worked on many articles where the established style is the first. In summary, on policy Oldsanfelipe, Trappist, Donald Albury and myself appear to be in agreement that page numbers should be used. Martin of Sheffield (talk) 17:24, 19 June 2018 (UTC)
"As matter of policy, it is uncontroversial that (aside from the lede text) every statement must have a source." No, that is not correct. An inline citation is only required if the statement is a direct quotation, has been challenged, or is likely to be challenged. Of course, it's often helpful to provide citations in other cases too.
Also, it is not required that successive statements each have their own citation if one location within one source supports several statements. The first example below overdoes citations, and the second example is more appropriate:
AD 2018 is year 6730 of the Julian Period.[1] The Jewish (A.M.) year 5778 begins September 20, 2018.[1]
AD 2018 is year 6730 of the Julian Period. The Jewish (A.M.) year 5778 begins September 20, 2018.[1]
Jc3s5h (talk) 17:27, 19 June 2018 (UTC)
Jc3s5h: how does your example contradict my statement, "every statement must have a source."?
"Every statement must have a source", as I read it, the editor must already have located a source for every statement, even if the editor does not cite the source in the article. The actual requirement from WP:V, "all content must be verifiable", means the editor should be confident a source can easily be found in the unlikely event the statement is challenged. So I'm allowed to add "when an NPN bipolar junction transistor is operating in the active region, current flows out of the emitter", because I know that information can be found in thousands of sources, even though I don't have one of the sources on the couch next to me opened to the appropriate page. Jc3s5h (talk) 20:16, 19 June 2018 (UTC)
And someone may come along and think that it needs to be cited, and tag it with {{cn}}, at which point it needs to be cited because it has been challenged and is not "common knowledge". · · · Peter (Southwood) (talk): 07:59, 21 June 2018 (UTC)
From my editing experience in the past, you can have a book open in front of you while editing, and even place a multi-page citation at the end of a paragraph, but you may have trouble finding which page in the source supports a specific statement in the paragraph when that statement gets challenged years later. Without even a clue as to what book or other source was consulted in writing the article, it can be very difficult to recover a source. What you think is obvious or well-known is not so for many readers. And what some people think is "common knowledge" turns out to be not-true. So, policy may only require that something added to an article be verifiable, but it saves time and trouble later on if everything added to Wikipedia is cited to a reliable source. - Donald Albury 10:48, 21 June 2018 (UTC)

────────────────────────────────────────────────────────────────────────────────────────────────────If the editor doesn't know anything about the topic, and had to look up every tidbit in a source, then Albury's comment is true. But if the editor is a subject matter expert, having to look up each fact, no matter how well known, before adding it makes editing much more difficult. Having that as a policy would change the Wikipedia policy from "the encyclopedia anyone can edit" to "the encyclopedia only ignorant people can edit". Jc3s5h (talk) 11:24, 21 June 2018 (UTC)

That doesn't track. Being an expert who knows whether a fact is real, from professional experience, has nothing to do with whether they need to prove that it's real; in fact, the expert is far more likely to be able to do this well, already being familiar with standard reference works in the field in question, and is more likely to either have them at hand or to be able to get to them easily (e.g. through journal-site paywalls). Being broadly ignorant of the topic means you're less likely to a) have any idea whether the claim is legit, or b) have any idea who to prove it. More to the point, there's no connection between "I know my stuff" and how to write encyclopedic material. The end reader has no idea who wrote which words in the article, and even if they dug that out of the page history, they have no reason to know that editor IPFrehley has a doctorate in the field while the next editor at the page Jimbo69, is a 16-year-old stock clerk at Wal-Mart who doesn't even have GED. I've had direct experience with alleged subject-matter experts being massively problematic at articles like Albinism, just deleting sourced material and inserting unreferenced claims based on what they learned in med school 30 years ago rather than based on what current sources know about the condition and its causes. It's the old "I'm an expert so I am a reliable source" confusion; WP relies on published sources, not claims about what is inside one's grey matter.  — SMcCandlish ¢ 😼  11:49, 21 June 2018 (UTC)
If the original editor is not around when a statement is challenged, then someone else has to come up with the citation, or the statement may be removed. Not every editor has sources on the tip of their tongue. I tend to edit articles relating to history and archaeology. While I (think) I know facts about a historical subject, I do not have a photographic memory, so I have to search to find a reliable source that supports what I remember reading in the past. Even if I have a suitable source at home, it still takes time to find what I want. If I originally found something in a library book, and didn't note the page number at the time, it is even more trouble to track down the reference. And, every once in a while, I can't find the source, and have to let something be deleted. I am an amateur historian, so I have to depend on reliable sources. I have a PhD in linguistics, and worked many years with computers, and I do not edit Wikipedia articles in those fields. Too much temptation there to rely on what I know (which is likely out of date) rather than what I can find in reliable sources. - Donald Albury 12:00, 21 June 2018 (UTC)
There is nothing wrong with providing more references than strictly needed as convenience service to readers. However our policies don't really require that and for good reasons. Individual cases need to be approached with some measure of common sense rather than absolute rules. What "common knowledge" is and how individual sourcing is minimally required depends on the article's topic and the primary target audiences. "Common knowledge" with regard to target audiences are in doubt well known facts/domain knowledge, that can easily be looked up in (any) standard textbook in the issue or reference books. None of that however justifies the "I add that because i'm an expert"-approach, but the justification is always that it can indeed be looked up easily. The latter than also can be used to be added as a source if really needed.
As far as unsourced sentences are concerned, simply removing them just because one can't find a source (quickly) is not always a good idea, but this heavily depends on context as well. In articles which are apparently well written and informed, one should remove unsourced content only if there is at least some (additional) reason to be believe the statement could be wrong, seems questionable or might violate policy. Simply not being sourced and not being able to find a source right away is often not good enough to justify a removal. The last thing we need is a "quality assurance" that combs through articles the topics of which it has no clue about and starts removing content it deems unsourced and for which it can't quickly find (not to say google) one.--Kmhkmh (talk) 12:40, 21 June 2018 (UTC)

Subject matter experts should provide citations for anything that is the least bit obscure or difficult to look up. But, in an article full of calculus and differential equations, requiring a subject matter expert to provide a citation to show that a quotient is the result of a division is just plain hostile to the editor, and to the readers who are competent to read the article because those readers have to wade through lots of unnecessary citations. Jc3s5h (talk) 13:12, 21 June 2018 (UTC)

References

  1. ^ a b c Astronomical Almanac for the year 2017, (Washington: US Government Printing Office, 2016), p. B4

Bundling cites and CITEVAR?Edit

Does switching to bundled citations count as a change which requires consensus, per WP:CITEVAR?

If so, should this be made clearer at CITEVAR?

@Emeraude:

See Britain First, [1][2][3] Andy Dingley (talk) 09:34, 20 June 2018 (UTC)

Also at Freemen on the land Andy Dingley (talk) 15:37, 20 June 2018 (UTC)
I would argue not, if the nature of the reference remains fundamentally the same. In those diffs the content of the citation templates have remained the same, so the citation style has remained the same. If for example an article uses Harvard referencing the guideline is supposed to prevent an editor coming along and templating all the citations. What you've got there is a change in the footnoting format, but not a change in citation. However, if you believe that bundling is detrimental in some way you are under no obligation to accept the change. Betty Logan (talk) 10:30, 20 June 2018 (UTC)
It is a WP:CITEVAR thing, yes, but it can also be a tidying up thing. For instance, in the Diels-Alder reaction article, Ref 47 combines all 28 papers as a single ref. That's a tidying up thing because these papers aren't cited independently of one-another in this context and having 28 different [47][48]...[75] would be incredibly ugly/annoying (in print, this would simply be [47-75], but that's not feasible on Wikipedia). However, the clustering of references [1][2][3] in the lead would be a clearly editor-and-reader hostile downgrade because they're cited independently of one another and re-used in the article in different places, for different things.
I have no opinion on if those diffs are tidying or not, but in my experience, bundling for bundling's sake is bad practice. Headbomb {t · c · p · b} 10:43, 20 June 2018 (UTC)
I don't believe is necessary, nor even sufficient, to cite 28 papers to establish that Deils and Alder wrote 28 papers about that reaction; that is more of a matter for a secondary reference. It is an anomalous case that does not show a general value of bundling; a better "tidying up" would be to replace that list of papers with a single, appropriate citation. ♦ J. Johnson (JJ) (talk) 20:22, 20 June 2018 (UTC)
I believe it falls under CITEVAR because that applies not only to the rendered appearance of the citation, but the approach future editors will have to take when editing the article. Bundling makes it a little more difficult to reuse or change citations. This is much the same as changing to list defined refereces, which also primarily affects how future editors edit the article. I also agree with Headbomb that bundling for bundling's sake is not very helpful. The situation where I would be likely to bundle would be if the footnote explains the relationship among sources,for example, "Brown[1] uses the data of Jones[2] and the curve fitting of Young[3]". Jc3s5h (talk) 12:44, 20 June 2018 (UTC)
Or something like

[1] Original ref

            Errata

. Headbomb {t · c · p · b} 13:37, 20 June 2018 (UTC)

Thanks for the comments. CITEVAR has never been clear over whether "technical" changes, such as whitespace (and linebreaks) around template params, which do not change the rendered content are within its scope or not. Although a switch to listing named refs within reflist (very innocuous, IMHO) does seem to be regarded as contentious, and the use or not of templates is unsurprisingly so. Where the results of the citations change though (and this is surely the case for bundling), then isn't that precisely why we have CITEVAR, thus (as here) they're covered by it?
As to most of the situations, such as that described by Jc3s5h, I would regard that "bundle" as a footnote, not a citation. I'd mark it up as a footnote, then have the three separate citations from within that. For the Diels-Alder case I might go further than that and set it as a paragraph of text or list: it's pretty much a historical bibliography of one author's work. My main concern with bundled citations is that they make it impossible to share citations to the same reference. Andy Dingley (talk) 15:37, 20 June 2018 (UTC)
The key points here are "it can also be a tidying up thing" and "bundling for bundling's sake is bad practice" and "Bundling makes it a little more difficult to reuse or change citations". I.e., this is not a citation "style" matter, it's WP:Common sense and practicality one. Sometimes bundling is the best idea, often it's not, and that rarely if every has anything to do with the citation style chosen, but rather is about how the specific sources are being used. The danger in trying to add this to the increasingly fetishized CITEVAR is that if someone does really bad bundling, we're likely to be stuck with it. (Or to be stuck with never doing bundling even when it makes sense, because some territorial twit will claim that doing the sensible thing is a violation of their precious CITEVAR natural rights. So, just don't.
Re, 'CITEVAR has never been clear over whether "technical" changes, such as whitespace (and linebreaks) around template params, which do not change the rendered content are within its scope or not.': We had an RfC on this a year or two ago, and the result was "no, they're not". When it comes to spaces, just (per WP:Common_sense again) do what everyone else is doing which is {{cite foo |title=Blah blah |first=Same |last=Bazzquxx |date=2018 ...}} This has been the vast-majority style for over a decade. It's simple, it's readable, it doesn't waste space, and it keeps each entire parameter grouped as a unit. More common sense: Vertical citations make sense at the bottom of the article (e.g. in WP:LDR layout); they do not work well in mid-article, because they make it hard to clearly discern the paragraph structure of the material. Thus virtually no one puts vertical citations in mid-article, meanwhile most LDR material is done vertically, where the cite details are easier to read. Context matters. In the context of prose, inline cites work better by not screwing with the ability of editors to make sense of the material without headaches. In the context of a thick pile of cite after cite, separated from the prose, the vertical cites (just like infobox parameters at the top of the article), are easier to manage. The more people a) apply common sense, and b) do what the consensus of other editors are doing instead of doing something weird because someone think they're unique snowflake, and c) stop bible-thumping CITEVAR like some kind of 11th Commandment from God, the better off we all are.  — SMcCandlish ¢ 😼  19:24, 20 June 2018 (UTC)
On the basis of changed appearance of rendered content: "bundling" (in its various forms) is not a small matter of "tidying up". It is a definite and significant change of appearance, certainly greater than the difference of using periods or commas for field separators (a key difference of CS1 versus CS2). On this basis alone CITEVAR is applicable.
Nonetheless, the objections to bundling go beyond appearance. I agree with Headbomb that "bundling for bundling's sake is bad practice." This in part because (as Jc3s5h and Andy say) bundling makes it more difficult to "reuse", share, or change citations (which I think applies to all use of "named-refs"). It is also confusing to the readers when a list of "references" (bad term) contains scattered sublists.
I disagree with Andy that "listing named refs within reflist" is "very innocous". But I strongly agree that there needs to be a stronger distinction between citations and footnotes. A footnote – or simply note – is the place or container created with <ref>...</ref> tags. Which may contain individual "citations" (full or short), bundles of citations (full or short), comments, or combinations thereof.
It should also be noted that questions of "bundling" seem to always involve full citations. Bundling of short-cites – rather than stringing them out in a series of notes – is actually preferable, and I recall no cases where that has been contentious. The recurring problem is not simply "bundling", it is bundling of full citations within notes. ♦ J. Johnson (JJ) (talk) 20:29, 20 June 2018 (UTC)
CITEVAR is very applicable here - articles that do not use bundled citations should not be switched to use them without a clear justification, and ideally not without discussion ahead of time to establish consensus. "One reference in each footnote" is a perfectly acceptable style for an article, and so this edit should have had a talk page discussion first. "Tidying" is often a synonym for "personal preference", which is not a valid reason to change established citation styles. — Carl (CBM · talk) 19:19, 21 June 2018 (UTC)
I'd say it's fine to be WP:BOLD in most cases of WP:CITEVAR, but WP:1RR applies when there's pushback. Headbomb {t · c · p · b} 19:27, 21 June 2018 (UTC)
I'd take a more conservative approach. This is the same principle as with ENGVAR - it doesn't say "it's OK to change from British to American English as long as nobody complains" - even if you think you are just tidying up. The language of both MOS passages is pretty clear against bold changes: "If the article you are editing is already using a particular citation style, you should follow it; if you believe it is inappropriate for the needs of the article, seek consensus for a change on the talk page." — Carl (CBM · talk) 19:51, 21 June 2018 (UTC)
The difference mostly being that "colour vs color" doesn't yield to any improvements on enhanced in functionality, whereas citation style is often shit, sub-optimal, or even completely un-established with an incoherent mishmash of styles. I've changed citation styles on hundreds of articles without any fuss, but they were also smart changes done for a reason, rather than simply personal preferences. (E.g. converting a {{harvnb}} to a non-harvnb system would not yield any significant improvements, but changing manual citations to templated ones will yield huge improvements in consistency, completeness, and functionality.) Headbomb {t · c · p · b} 20:32, 21 June 2018 (UTC)
If there was a broad consensus that a particular change would be an improvement, we would put it in WP:CITE. The reason we don't have a preferred style is that there is no consensus that adding templates, bundling citations, or many other things is actually generally an improvement. In the end, it's all just personal opinion, like "colour" vs. "color". That's why we use the "go with the first established style" policy, because there is no broad consensus about which method to use. — Carl (CBM · talk) 21:06, 21 June 2018 (UTC)
Yes. Dropping a line on the Talk page asking if anyone objects is so little trouble, and can so easily avoid a big battle and possibly having to scrap a significant investment of time and effort, that anyone failing to do so ought to spend some quiet time in a corner.
Something else to consider: if CITEVAR does not apply to "bundling" (however conceived), then it likewise does not apply to unbundling. Whether a change is significant (on any basis), or not, applies in both directions. ♦ J. Johnson (JJ) (talk) 20:44, 21 June 2018 (UTC)
My experience matches that of CBM. I routinely impose a consistent citation system in the absence of one at an article, without any fuss whatsoever. The idea that WP is full of CITEVAR obsessives is just nonsense. It's really a total of about a dozen editors, and one simply learns to avoid them (and to not be a wikilawyer: if a drive-by editor added a differently formatted citation to an article that was consistently using another style until that point, that doesn't magically make the article "inconsistent" in a way that makes it okay to subject the entire thing to the newly introduced style).  — SMcCandlish ¢ 😼  05:40, 22 June 2018 (UTC)
I have a preferred style for citations that I use on new articles and when expanding un-sourced or poorly sourced articles. I don't remember anyone making wholesale changes to citations in an article where I have supplied most of the citations, but editors show up every once in a while who bundle or un-bundle citations, remove citations they consider to be excessive, or make other changes to individual citations. Unless such changes degrade the appearance of the article, or leave part of the article un-sourced, I generally ignore them. I don't like the stress of drama shows, so I try to avoid them as much as possible. - Donald Albury 12:36, 22 June 2018 (UTC)
Actually I'm pretty sure I've reverted changes by SMcCandlish, as I usually do with cite-bandits. There are quite a few of these, and they disrupt, sometimes totally prevent, further editing. I have had to abandon articles because siome fool has come along and imposed his own style, & I haven't noticed until there have been so many subsequent changes I can't be bothered to unpick them, or to struugle on with some ghastly style I don't understand. Typically, when it is pointed out that their behaviour is wholly against WP:CITEVAR, they are entirely unapologetic, and often arrogant with it. Johnbod (talk) 18:34, 23 June 2018 (UTC)

arbitrary breakEdit

  • The essence of CITEVAR is to seek consensus (if not explicit approval, at least a lack of objection) prior to making certain changes to citation. (The specific language is "without first seeking consensus".) It seems to me that CITEVAR spats are invariably cases where an editor claims some change is exempt from CITEVAR. I am increasingly convinced that (as I stated above) if some change is not subject to CITEVAR, then reverting such a change is also not subject to CITEVAR. Any issue is then one of WP:BRD. Which has led me to wonder: do those who claim exemption from CITEVAR also claim exemption from BRD?
In this light I am concerned about @Emeraude's assertion (in this edit summary) that "Agreement is unnecessary: see Wikipedia:Citing sources#Bundling citations". That section claims some advantages for bundling, but it does not say "agreement is unnecessary". Yet it appears some editors are claiming that exemption from CITEVAR's requirement to seek prior consensus means that no agreement is required, even in the face of an objection. That might be trivially true for an initial Bold edit, but in no way should be allowed as an exemption from the requirement for consensus where a change is contested.
Any confusion about this not only supports answering the question Andy posed in the affirmative, but also suggests the language of CITEVAR needs to made stronger: that all changes are subject to WP:BRD, and that notification should always be made before changing any aspect of citation (aside from adding citations in the established style). ♦ J. Johnson (JJ) (talk) 23:25, 22 June 2018 (UTC)

My feeling: if a change in how citations are handled (such as this bundling one) turns out to be controversial (as this one clearly is), then it is at least retroactively subject to CITEVAR: the change should be reverted and not reinstated unless/until consensus can be obtained. —David Eppstein (talk) 23:33, 22 June 2018 (UTC)

The problem with this reasoning, which we've been over many times before, is of course that actual policy doesn't agree, namely WP:EDITING and WP:OWN policies. No one needs prior approval/permission to edit any article. If they're not acting in good faith, or if what they do isn't constructive, anyone's free to revert, and the onus is on the one wanting to make the change to gain consensus that it's a good idea (while WP:BRD is technically an essay, the community treats it as more than that, the same way it approaches WP:AADD, WP:Common sense, and a few other "super-essays"). A habit of making bad faith or unconstructive changes is disruptive and addressable at noticeboards. The point of CITEVAR is to suggest as a guideline, a best practice, to avoid making unnecessary or potentially controversial changes to citations. This suggestion is made based on experience and predicability, so it is generally sensible, but as with all guidelines some exceptions may apply, especially if not applying them leads to excessive bureaucracy, like repetitive, time-sucking, pointless RfCs over trivial citation formatting details. CITEVAR is not Holy Writ or a law of nature, and it cannot be wielded against people like a weapon; trying to do so is itself disruptive. The goal is to have the best citation layout and coding for the article and its context. Territorially fighting half to death over how we get there is not working on the encyclopedia and is entirely missing the point of the endeavor.  — SMcCandlish ¢ 😼  12:48, 23 June 2018 (UTC)
Of course, the same applies to ENGVAR: it is not holy writ, but it does not violate WP:EDITING or WP:OWN to tell people not to change the style of English from American to British, and they cannot justify such edits by "I don't need permission to edit an article" or "I was jsut being bold" or "nobody complained". The same applies to CITEVAR and WP:STYLEVAR. The real difficulty with citations is when editors refuse to accept that no citation style is really better than any other, and instead insist on changing styles to the ones they personally prefer. If they would focus on more constructive improvements to articles, we would avoid a huge amount of wasted conversation, and the purpose of CITEVAR (like ENGVAR and STYLEVAR) is to be direct that these nonconstructive style, citation, or English variation changes should be avoided. — Carl (CBM · talk) 15:24, 23 June 2018 (UTC)
It's fine to make any kind of *VAR change, if the change is clearly an improvement. I wouldn't say I do it "routinely", since it doesn't come up that often, but I have numerous times changed the ENGVAR of an article to match the subject per MOS:TIES (e.g. an American biography written in British/Commonwealth English, or vice versa, more often vice versa due to the number of American editors). I don't go beg for permission first, I just do it when it's the right thing to do, and no one fights me on it. I'm not making any kind of argument about CITEVAR that I wouldn't make about ENGVAR or DATEVAR or TITLEVAR. It actually most often comes up in DATEVAR cases, because people use citation scripts indiscriminately, with hard-coded date formats, and they just DGaF what the article's established (or common-sensical) date format might be. "If they would focus on more constructive improvements" really is the key. But it's also central the problem with "CITEVAR obsessiveness" or whatever one might want to call it; the territorial response can be so thickheaded that change will be resisted simply because it's change. The reason certain individuals like and rely on CITEVAR isn't what's best for the article, it's what's best for their personal preferences and sense of control. We don't have that problem with any of the other *VARs. It's why things like this keep coming up here.  — SMcCandlish ¢ 😼  18:18, 23 June 2018 (UTC)
Well, WP:ENGVAR really has no tradeoff. colour vs color has no benefits, although I fully agree with SMcCandlish that bringing an article in line with WP:MOSTIES doesn't need begging for permission first. WP:CITEVAR however, has plenty of WP:IAR situations. For instance, one could write an article in a completely consistent, but fully non-standard "AUTHOR. YEAR, TITLE. DAY-MONTH, PAGE. VOLUME, JOURNAL: ISSUE.". This is a scheme that makes zero sense, and pretty much no one would get in trouble for boldy converting that to a proper {{cite xxx}} scheme. Where WP:CITEVAR applies is in converting something like Quark#References into a {{sfn}}/{{harvnb}} scheme, or going from CS1 style to CS2 style. Headbomb {t · c · p · b} 18:38, 23 June 2018 (UTC)
Even if someone think an article has an idiosyncratic style, it's still not in accordance with CITEVAR to boldy change it: "if you believe it is inappropriate for the needs of the article, seek consensus for a change on the talk page". In the case mentioned, it would really make little sense to take an article with a consistent non-templated system and replace it with a templated system only because the non-templated system seemed idiosyncratic - CITEVAR specifically discourages "adding citation templates to an article that already uses a consistent system without templates". There is nothing particularly "proper" about templates, which are neither preferred nor discouraged as a method for citations. — Carl (CBM · talk) 18:46, 23 June 2018 (UTC)
And that's where WP:BURO/WP:IAR applies. If a rule prevents you from improving Wikipedia, fuck the rule. And that is policy. That said, if you get pushback on this, it's best to follow a WP:1RR mindset, mostly per WP:CITEVAR. Headbomb {t · c · p · b} 18:49, 23 June 2018 (UTC)
The point of CITEVAR, ENGVAR, etc. is that we have a consensus that particular kinds of changes generally do not improve Wikipedia, making IAR more difficult to apply :). — Carl (CBM · talk) 18:51, 23 June 2018 (UTC)
Experience and reality disagrees with that. I've made zillions of those changes with little to no pushback, and will keep doing so. Headbomb {t · c · p · b} 19:02, 23 June 2018 (UTC)
The very fact that someone will argue against the idea of changing without prior consensus discussion away from an insane and confusing, totally made-up, idiosyncratic citation "style" used by no one else in the world, such as illustrated by Headbomb, is the very crux of CITEVAR's long-running WP:CONLEVEL problem. No one on WP believes this other than a handful of people who spend too much time on this talk page. It's turning into a deeply un-wiki echo chamber. We should probably just to RfC the matter at WP:VPPOL. I would happily place a large wager on the outcome, since the policy problems with this idea are really quite clear. PS: The idea that "template[d citations] ... are neither preferred nor discouraged" is clearly not really true. Actual usage proves it, and that the guideline text is divergent from actual, operational consensus, as it has been for a very long time. I'm not here to try to change it, since I find drama tiresome, but WP:P&G instructs us to have guidelines codify practice not try to dictate it.  — SMcCandlish ¢ 😼  19:47, 23 June 2018 (UTC)
@SMcCandlish - "it's what's best for their personal preferences and sense of control" applies equally well to the editors who change the established system at an article, when in reality no system is better than any other one. If a particular change is an improvement, it will be documented in the MOS - as with MOS:TIES. For changes covered by one of the VAR policies, if a particular change has consensus as an improvement, it will be listed in the MOS, So, if a change is not required by the MOS, we should assume there is no consensus that the change is actually an improvement, or develop consensus to change the MOS first. — Carl (CBM · talk) 18:51, 23 June 2018 (UTC)
Sure, many arbitrary changes to cite style really are just arbitrary, like insisting on Vancouver format, or whatever. But many of them are not. Switching from unformatted citations to CS1, for example, produces objective benefits. Depending on the nature and layout and stability of the article, so can moving to WP:LDR (e.g. at and above the WP:GA level) or away from it (at the stub to C-class level). Harvard referencing and SFN are very helpful in some cases, but onerous impediments for the average editor in others. Horizontal cites work better in prose, vertical ones in LDR. The supposition that a particular citation style and/or formatting scheme is always better or worse regardless of context is like assuming that a reliable, secondary source for something is both reliable and secondary for everything. We actually have long-term editors who do not understand why those assumptions are false. I'm sure we also have editors who insist that their preferred cite style is the One True Style, but they're cranks and we can ignore them.

"So, if a change is not required by the MOS, we should assume there is no consensus that the change is actually an improvement" – except not, per WP:EDITING, WP:AGF, etc. WP's actual assumption is that any change an editor wants to make is probably an improvement, and no one is barred from trying to make one. If that assumption turns out to be untrue or questionable in a particular case, revert away. CITEVAR suggests that the assumption is less likely to be true for citation style changes than for most other kinds of editorial alterations, but that doesn't mean that the assumption a citation change isn't an improvement is automatically correct. Otherwise we would not have a CITEVAR guideline, we would have a "WP:NOCITECHANGE" policy that flatly forbade any citation formatting changes without a pre-established consensus. Never going to happen.

Your formulation is a pretty good summary, if you shift the emphasis point: We have a consensus (in the *VAR guidelines) that particular kinds of changes generally do not improve Wikipedia, making IAR more difficult to apply. More accurately, we have consensus to warn that those kinds of changes are more frequently controversial and potentially disruptive. "Generally" is not synonymous with "always". Experienced editors who are not flaming asshats usually learn to figure out competently where "generally" does/doesn't apply. If they didn't, we wouldn't have any "generally" rules of any kind, only legalistic absolute ones, and no IAR policy at all. The problem we have is that a handful of CITEVAR aficionados deny these realities and treat CITEVAR as if it were the imaginary "NOCITECHANGE policy". Yes, there's a countervailing problem of random drivebys changing citation formats completely for no good reason. I'm sure that's frustrating, but it's easy enough to revert it, and in the end the readers don't care, and 99+% of editors don't either. It's easy to learn to care less and to just absorb multiple citation systems; if I have, you all can too. (Ages ago I was opposed to CITEVAR existing at all, and was in favor of CS1 being the only recognized WP citation format.)
 — SMcCandlish ¢ 😼  19:47, 23 June 2018 (UTC)

Return to the project page "Citing sources".