Wikipedia talk:Citing sources

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

This is an old revision of this page, as edited by Bagumba (talk | contribs) at 16:00, 14 September 2023 (→‎WHENNOTCITE vs DYK: clarify). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

Wiki DOI Gbooks Citation Maker is Down

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)[reply]

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)[reply]

General references

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)[reply]

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

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)[reply]
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)[reply]
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)[reply]
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)[reply]
@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)[reply]
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)[reply]
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)[reply]
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)[reply]
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)[reply]

Pages in PDF

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)[reply]

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)[reply]
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)[reply]

References

  1. ^ John R. Ehrman (February 2016). Assembler Language Programming for IBM System z ™ Servers - Version 2.00 (PDF) (Second ed.). IBM Silicon Valley Lab. p. 4. Retrieved July 12, 2023. 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.

Citing your own book, article, working paper

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)[reply]

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)[reply]
"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)[reply]
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)[reply]

Random not-an-admonishment formatting

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)[reply]

Update guide to reflect deprecation of parenthetical referencing?

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)[reply]

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

Proposed change

Replace:

With:

Reason:

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

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)[reply]

Adding a quote to a ref name

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)[reply]

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

WP:REPEATCITE

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)[reply]

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)[reply]
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)[reply]
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)[reply]
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)[reply]
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)[reply]
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)[reply]
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)[reply]
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)[reply]
@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)[reply]
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)[reply]

Reference order

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)[reply]

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)[reply]
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)[reply]
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)[reply]
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)[reply]
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)[reply]
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)[reply]
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)[reply]
Yep. Stuff is apt to move around a lot.  — SMcCandlish ☏ ¢ 😼  20:27, 27 August 2023 (UTC)[reply]
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)[reply]
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)[reply]
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)[reply]
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)[reply]
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)[reply]
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)[reply]
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)[reply]
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)[reply]
Opened the RfD.  — SMcCandlish ☏ ¢ 😼  20:52, 27 August 2023 (UTC)[reply]
Shouldn't that read "numerical" rather than "chronological"? -- Valjean (talk) (PING me) 20:47, 27 August 2023 (UTC)[reply]
Yes, good point. Will tweak it in a sec.  — SMcCandlish ☏ ¢ 😼  20:52, 27 August 2023 (UTC)[reply]
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)[reply]
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)[reply]
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)[reply]
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)[reply]
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)[reply]
'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)[reply]
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)[reply]
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)[reply]
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)[reply]
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)[reply]
Thanks -- that's a helpful explanation. Mike Christie (talk - contribs - library) 22:02, 27 August 2023 (UTC)[reply]
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)[reply]

Proposed wording change re in-text attribution to conform to V and NONFREE

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)[reply]

  • 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)[reply]
  • 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)[reply]
  • Support per the rationale of the three above editors. - SchroCat (talk) 19:12, 27 August 2023 (UTC)[reply]
  • 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)[reply]
  • 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)[reply]
    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)[reply]
    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)[reply]
    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)[reply]
  • Support because it's confusing to have slightly different policies on different pages, Rjjiii (talk) 19:44, 27 August 2023 (UTC)[reply]

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)[reply]

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

WHENNOTCITE vs DYK

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)[reply]

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)[reply]
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)[reply]
@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)[reply]
Whoops, sorry, didn't realize I was replying to a discussion that was 9 days old! Valereee (talk) 12:34, 9 September 2023 (UTC)[reply]
@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)[reply]
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)[reply]
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)[reply]
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)[reply]
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)[reply]
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)[reply]
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)[reply]
I've added a footnote, anyone should feel free to correct whatever I did wrong! Valereee (talk) 15:36, 9 September 2023 (UTC)[reply]
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)[reply]
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)[reply]
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)[reply]
...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)[reply]
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)[reply]

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)[reply]

References

  1. ^ a b c Citation for sentence with hook and subsequent sentence(s)
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Citing_sources&oldid=1175370901"