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 nohouse style on Wikipedia, and the community does not want to have theholy 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 shouldnever edit-war over the style of citations.
What styles are commonly used?
There aremany published style manuals. For British English theOxford Style Manual is the authoritative source. For American English theChicago Manual of Style is commonly used by historians and in the fine arts. Other US style guides includeAPA style which is used by sociologists and psychologists, andThe MLA Style Manual which is used in humanities. TheCouncil of Science Editors andVancouver 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] to cite sources in an article?
Not technically.Footnotes (<ref>...</ref>) are by far the most popular method of placing citations. A few older articles may still use anow-deprecated form ofinline parenthetical referencing. Manual systems, usually using symbols† or lettersA, are used in a few articles (example). A very short article might not be required to use inline citations. SeeWikipedia:Inline citation for more information.
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.
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 arepublished and accessible to the public.
To find archives of this talk page, seethis list. For talk archives from the previous Manual of Style (footnotes) page seeHelp talk:Footnotes.
This project page does not require a rating on Wikipedia'scontent assessment scale. It is of interest to the followingWikiProjects:
This page falls within the scope of theWikipedia:Manual of Style, a collaborative effort focused on enhancing clarity, consistency, and cohesiveness across theManual of Style (MoS) guidelines by addressing inconsistencies, refining language, and integrating guidance effectively.Manual of StyleWikipedia:WikiProject Manual of StyleTemplate:WikiProject Manual of StyleManual of Style
This page may fall under thecontentious topics procedure and be given additional attention, as it may be closely associated to thearticle titles policy andcapitalisation. Both areas are subjects of debate. Contributors are urged to review theawareness criteria carefully and exercise caution when editing.
This page is within the scope of theWikipedia Help Project, a collaborative effort to improve Wikipedia's help documentation for readers and contributors. If you would like to participate, please visitthe project page, where you can join thediscussion and see a list of open tasks. To browse help related resources see theHelp Menu orHelp Directory. Orask for help on your talk page and a volunteer will visit you there.Wikipedia HelpWikipedia:Help ProjectTemplate:Wikipedia Help ProjectHelp
Hi all! Relatively few newcomers know, let alone remember, to cite the content they are adding. In response, the WMF's Editing team has been working on developingReference Check, a feature which prompts new editors to add citations before they publish an edit adding content to an article. We're hoping to bring it to English Wikipedia soon with an initial A/B test so that we can evaluate its impact and address any issues that arise. As the first stage of that, we wanted to reach out to gather your input on the feature and experiment's design.
Background
Reference Check is the first of a new set of features calledEdit Checks, which aim to:
Give newcomers guidance while they are in the process of making an edit to help them abide by Wikipedia's policies and guidelines.
Reduce the amount of effort and attention experienced editors need to allocate toward addressing preventable damage.
In this case, it aims to address the problem that only 19% of edits adding content by new editors (those with ≤100 edits) include a reference. It has been in development since 2023, and isnow deployed on every other language edition except for English.
How it works
Reference Check
Reference Check activates when a new editor adds a large amount of text in VisualEditor (on desktop or mobile) in an edit and clicks on publish without adding a citation. It creates a notice (see screenshot) prompting the editor to add a citation. If they choose to do so, it opens up the "Add a citation" box, and if they decline, it asks them to specify why, which is then recorded (and weintend will eventually be available for other editors to review).
Our research from other languages shows a positive impact: Inan A/B test, editors who were shown the check were more than twice as likely to make an edit that included a reference, as well as less likely to be reverted and more likely to stick around and keep editing.
There have beenseveralpriordiscussions where some editors have expressed interest in deploying it here, and we're now looking to follow up on those and complete the deployment in the near future.
Create a new paragraph that is at least 50 characters longwithout adding a citation.
Press the "Publish changes…" button.
Interact with the prompt that appears.
Be sure not to click "Publish changes" (without the...) from the "Save your changes" screen so that you don't actually publish the test edit.
Call for feedback
We’re hoping to runan A/B test showing the feature to a sample of newcomers who have made ≤100 edits so that we can evaluate its impact and address any issues that arise. We're interested to know what you think of this approach. Overall, does the check seem like something that would benefit the project? Is there anything you'd want us to look for or keep in mind during the test? (The current metrics we plan to track aredocumented on Phabricator.) We also continue to be open to more general feedback if there's anything else that testing it out brings to mind.
We'd also like to highlight that there areCommunity Configuration settings available for Reference Check, which allow any admin to adjust things like how many characters need to be added for the check to activate. Feel free to discuss existing settings options that you'd prefer, or to let us know if there are additional settings you'd like us to introduce.
We'll follow up after gathering input with next steps. And thanks as always for your collaboration!
It looks for one citation per paragraph, which is perhaps the best that can be easily managed and certainly better than nothing, but will often be inadequate.
The four options offered when you decline to add a citation are insufficient. In particular, expanding from a preexisting source is quite common. Ideally when a new paragraph is used a NAMEDREF should be repeated, but newcomers may not expect that and an option that provides the opportunity to easily reuse an existing reference after an initialno is worth considering.
Some articles will not have any lead citations, bit of an edge case and maybe not worth the bother of an extra option but rewriting of leads is sometimes done by new users, especially if the existing one is self-evidently inadequate.
Theother option offers no chance for further explanation which might be useful if for no other reason the data collection.
Thanks for sharing these thoughts; they are all very helpful!
For making it easier to reuse references, we'll discuss this and follow up with any updates.
For articles without lead citations, one of theCommunity Configuration options,ignoreLeadSection, deactivates the check there.
For theother option, the team had initially decided not to prioritize that because of the complexity it'd introduce, but it remains a possible future enhancement. @PPelberg (WMF) has createdphab:T405683 to document that work.
Probably also pages tagged as a set index, as for example many chemical compounds are 50 characters long even without an accompanying description. Also applies equally to certain other pages that are purely navigational like lists of lists though implementation may be too difficult in practice and users are allowed to selectno when queried as whether a reference is needed so maybe not that big a deal though data is limited.184.152.65.118 (talk)22:28, 25 September 2025 (UTC)[reply]
This seems like a very useful tool, thank you to all those involved in making it. Out of curiosity, what does it look like when an editor chooses not to add a reference? Is the edit tagged with the reason they gave?Toadspike[Talk]22:33, 26 September 2025 (UTC)[reply]
Great questions, @Toadspike. Responses below. Please let me know what (if any) new questions this brings to mind or leaves unanswered...
...what does it look like when an editor chooses not to add a reference?
When someone electsnot to add a reference, Reference Check will ask them to express why (see screenshot).
Reference Check mobile decline survey
Is the edit tagged with the reason they gave?
At present, no. The reason someone gave isnot visible on-wiki. Although, we are consideringre-introducing[1] this functionality in the future viaT405132.
One of the recurring concerns is to what share of new contributions are improved vs what share are introducing more subtle mistakes. I dont expect too many problematic subtle mistakes here, but getting a list of edits in the two groups and allowing the community to compare might be nice.—Femke 🐦 (talk)16:53, 15 October 2025 (UTC)[reply]
...getting a list of edits in the two groups and allowing the community to compare might be nice.
Good call. And you know what, Ithink generating the lists of edits you described above will be possible using theEdit Check edit tags that are already in place...
editcheck-references: ought to return a list of edits that meet the conditions for Reference Check to be shown had it been enabled
editcheck-references-shown: ought to return a list of edits in which Reference Check wasactually shown to someone in the course of publishing an edit
@DLynch (WMF): can you please confirm the above is accurate? And assuming David confirms the above to be true, @Femke do these two tags sounds like they'll offer the info. you think would be helpful for volunteers to see?
Yes, that looks correct. (And thus, effectively,editcheck-references +editcheck-references-shown on thesame revision should mean a revision where someone was asked to add a reference and chose not to.)DLynch (WMF) (talk)17:15, 15 October 2025 (UTC)[reply]
Hi all! Based on the discussion above, we're planning to move ahead with the A/B test. The test is scheduled to run from Wednesday, 5 November 2025 to approximately 17 December 2025, after which we'll take some time to analyze the results, share them with you all, and decide with you how/if to enable the feature. Thanks for all your feedback above! We'll continue to monitor this thread, so if anything else comes to mind or you notice anything during the test, please let us know! Cheers,Sdkb‑WMFtalk21:31, 3 November 2025 (UTC)[reply]
@Thryduulf,Femke, andToadspike: Those settings (or whatever other ones you decide to adopt — the community configuration settings are yours to customize how you see fit!) are all fine by us. Feel free to implement them at will on the MediaWiki page. Cheers,Sdkb‑WMFtalk15:41, 5 November 2025 (UTC)[reply]
And like I said in that suggestion on the message's talk page, I'm happy to make the actual edit for you if you don't feel comfortable adding the JSON yourself -- I just don't want to step on anyone's toes by unilaterally running ahead with that.DLynch (WMF) (talk)22:10, 5 November 2025 (UTC)[reply]
It should hopefully not show up on those pages, since (per Ed above) it doesn't trigger on list items. As a backup, we are exploring ways to suppress the feature on pages based on category (seephab:T347775), but that hasn't been implemented yet. Cheers,Sdkb‑WMFtalk22:16, 6 November 2025 (UTC)[reply]
Somewhere there must be a bit of software that generates{{cite news}} citations from newspapers online, but it has a weakness. Here's a typical example (fromOperation Prone).
{{Cite news|url=https://www.nytimes.com/1988/08/09/world/aug-20-cease-fire-8-year-iran-iraq-war-southern-africa-pact-set-too-angola-truce.html|title=AUG. 20 CEASE-FIRE IS ON IN 8-YEAR IRAN-IRAQ WAR: SOUTHERN AFRICA PACT SET, TOO; Angola Truce Now|last=Times|first=Robert Pear, Special To The New York|date=9 August 1988|work=The New York Times|access-date=2017-07-11|language=en-US|issn=0362-4331}}</ref>
It looks to me as though it sees ", Special to the New York Times" in the 'author' slot of the source, and assumes from the comma that it's in (last, first) format, and then makes "Times" the last name and everything preceding that the first name (this clearly isn't a fully correct explanation though). There are many similar examples (not just for the NYT). Does anyone know anything about this? Is it just on older citations, or are they still being generated? Can we do anything about it?Colonies Chris (talk)10:13, 26 September 2025 (UTC)[reply]
To be clear, the "Special to The New York Times" part shouldn't be in the 'author' tag of the original article, and isn't for newer NYTimes articles, but older articles are often shoved into a more modern straight jacket that doesn't account for all the variance in the history of the paper's publishing. —TheDJ (talk •contribs)09:08, 1 October 2025 (UTC)[reply]
What software is this - is it something written for Wikipedia, or provided by an archival website? If it's ours, could it be tweaked to spot the phrase ", Special to xxxxx" at the end of the 'author' and deal with it more appropriately when building the citation?Colonies Chris (talk)09:15, 1 October 2025 (UTC)[reply]
@Colonies Chris, good question. I believe the tool that generates the citations ismw:Citoid, which in turn usesZotero to "translate" website metadata from a given URL into the appropriate fields.
Since Zotero is open-source, I believe the way to address this issue would be to go to its platform and try to fix it there.Sdkbtalk22:22, 5 October 2025 (UTC)[reply]
Taking a broader view, I think we do a terrible job of surfacing the fact that this is what's happening under the hood, which in turn makes it harder for people to report issues like the one you found. I'd be interested to hear from you and any editors with experience with Zotero: Are there places we could add this info? I'm thinking of something like a small notice in the editor after it autogenerates a citation with "Report a problem with this citation's formatting" that'd lead to somewhere that'd let editors work on fixing it.Sdkbtalk22:23, 5 October 2025 (UTC)[reply]
The problem is in the way that news outlet chooses to set up their HTML (the 'author' tag). Here's what any tool is starting with for that article:<metadata-rh="true"name="byl"content="By Robert Pear, Special To the New York Times"/><ahref="https://www.nytimes.com/by/robert-pear"class="last-byline css-ojhyzr e1jsehar0"itemProp="name">Robert Pear, Special To the New York Times</a>
With the website falsely claiming that this author has a name that's eight words long – and which a human, but not a bot, can see contains non-name information – there's not a lot that can be done by the tools, especially if you don't want to hand-code a thousand exceptions for all the different ways that all the different websites screw up their metadata.WhatamIdoing (talk)18:42, 11 October 2025 (UTC)[reply]
Yes, it's not practical to try to catch all exceptions, but there are some frequently encountered and easily identified patterns such as this one that could surely be fixed rather than propagated into our articles? It's common enough that there are hundreds of articles with citations such as the one I quoted.Colonies Chris (talk)10:58, 14 October 2025 (UTC)[reply]
Another common pattern is with items in the Google news archive, such as this one:
{{Cite web |title=Spokane Daily Chronicle - Google News Archive Search |url=https://news.google.com/newspapers?id=C5NYAAAAIBAJ&sjid=PvgDAAAAIBAJ&pg=6184,1800236&dq=long-island+golf+wright&hl=en |access-date=2022-11-23 |website=news.google.com}}
where the 'title' has been populated by the name of the publication, suffixed by 'Google News Archive', instead of the actual article title, which is nowhere to be seen, and the 'website' is somewhat misleadingly given as Google instead of the publication - and probably Google News Archive should be in the 'via' parameter.Colonies Chris (talk)11:34, 14 October 2025 (UTC)[reply]
@WhatamIdoing, as tedious as hand-coding exceptions to screwed up metadata is, I am with Chris in that it's still better than the alternative: hand-coding every citation to a site with screwed up metadata. And that latter task is much larger, since there will be e.g. hundreds of citations screwed up with the "Special to the New York Times" issue and ditto for every other issue. If we could easily point people to a system that'd allow them to write translators to fix these issues, and make that system easy enough to use, I think there would be sufficient editor interest to make it useful.Sdkbtalk17:52, 14 October 2025 (UTC)[reply]
That assumption has been proven false in the past. For themw:citoid service, the system isZotero, and anyone who wants to create, update, improve, etc. for any website at all is welcome to do so. Weeven provide instructions on how to do it. But I think I've only seen two editors do this in the last decade, and at least one of those had non-wiki experience with Zotero.
While fixing the Zotero translator is the Right™ way to do this, and would benefit not just us but all Wikipedias and even non-wiki people all over the world, given our skill sets, I think that if you want an automated way to handle a hand-curated list of known problems, we'd be better off asking for help 'after the fact' atWikipedia:Bot requests.WhatamIdoing (talk)18:04, 14 October 2025 (UTC)[reply]
ReThat assumption has been proven false in the past, I'd argue that the assumption has not been properly tested, since I included in my hypothesis that we'deasily point people to a system andmake that system easy enough to use. Right now, we are not easily pointing people to it (as evidenced by the difficulty Chris had in finding it despite being highly motivated) and we have not made it easy to use (as evidenced by the 5000-word "help" page you point to, as well as by the fact I looked into making a translator myself in the past and gave up after deciding it was too complex to be worth the effort).
Granted, the work of making Zotero easier to use is on the Zotero community, and there are probably some inherent complexities that limit how simple it could be made. But still, I don't think the reason people aren't making translators is that there's no interest in the task. Rather, it's that they don't even know it's a possible task in the first place, let alone how to do it.Sdkbtalk18:41, 14 October 2025 (UTC)[reply]
If there are no page numbers, whether inebooks or print materials, then you can use other means of identifying the relevant section of a lengthy work, such as the chapter number, the section title, or the specific entry.
Thanks for the ping! Imo it means "quoting part of the specific entry verbatim", but since we're here, we might as well discuss the actual matter instead of guessing what someone meant a few years ago.What should be another recommended way to cite a passage of material lacking page numbers (ebooks), besides the section title and chapter number?FaviFake (talk)17:47, 11 October 2025 (UTC)[reply]
Quotations are a valid way to indicate a place, but they are by no means the only way to indicate a place. These are "specific entries" that do not use quotations:
For a dictionary, encyclopedia, or other book with short pieces, give the name of the short bit:"example".Little's Little Dictionary.
For a codified law, give the code numbers: U.S. Const. art. III, § 2, cl. 3.
For a table or image in an unpaginated book: Give the title or number of the desired piece: "Table 3: Comparison of formatting techniques",TheStyle Book.
These examples are what we meant by "specific entries". It tells you where to look, using some method other than page numbers, chapters (which may be titles or numbers), section headings, or quotations.WhatamIdoing (talk)18:54, 11 October 2025 (UTC)[reply]
This sounds great! I think the second one is the same aschapter number, but the other ones are interesting. I think we should add them as additional examples.FaviFake (talk)19:04, 11 October 2025 (UTC)[reply]
I have also interpreted the older wording to mean anentry in a reference work. Quotations, "code numbers", the "title or number of the desired piece" or "the name of the short bit" are probably all effective alternatives when a page number is not available. For example:
Page number(s)
Stevenson, Angus, ed. (2010).Oxford Dictionary of English. Oxford University Press. pp. 1199–1200.ISBN978-0-19-957112-3.
Entry
Stevenson, Angus, ed. (2010). "Night".Oxford Dictionary of English. Oxford University Press.ISBN978-0-19-957112-3.
Quotation
Stevenson, Angus, ed. (2010).Oxford Dictionary of English. Oxford University Press.ISBN978-0-19-957112-3.night → noun 1 the period from sunset to sunrise in each twenty-four hours:a moonless night
"U.S. Const. art. III, § 2, cl. 3." is not "the name of the specific entry". AFAICTU.S. Const. art. III, § 2, cl. 3. doesn't have "a name". I think we should just go back to the original wording, and if necessary, link to something that defines "dictionary entry". If you want to be more vague, then consider "the relevant part of the source".
Your example, "U.S. Const. art. III, § 2, cl. 3." is an example of a chapter number. The chapter. number is entirely adequate to identify it. Here is the specific entry:
Representatives and direct Taxes shall be apportioned among the several States which may be included within this Union, according to their respective Numbers, which shall be determined by adding to the whole Number of free Persons, including those bound to Service for a Term of Years, and excluding Indians not taxed, three fifths of all other Persons. The actual Enumeration shall be made within three Years after the first Meeting of the Congress of the United States, and within every subsequent Term of ten Years, in such Manner as they shall by Law direct. The Number of Representatives shall not exceed one for every thirty Thousand, but each State shall have at Least one Representative; and until such enumeration shall be made, the State of New Hampshire shall be entitled to chuse three, Massachusetts eight, Rhode Island and Providence Plantations one, Connecticut five, New-York six, New Jersey four, Pennsylvania eight, Delaware one, Maryland six, Virginia ten, North Carolina five, South Carolina five, and Georgia three.
In terms of people understanding what's meant by "specific entry", I think that giving the example of a dictionary entry is probably the simplest. But we could just use other words. The point is to tell people that they need to use whatever is suitable and relevant for the source, even if the source does not have any chapter numbers or section titles (or if it has them but they are too large to be useful). If you can think of a better way to express this, then please feel free. What we don't want is for editors to think "Hooray, there's no chapter numbers or section titles, so I don't have to bother providing|at=details!"WhatamIdoing (talk)04:17, 12 October 2025 (UTC)[reply]
First to the point: "If you can think of a better way to express this, then please feel free. " I have proposed two different solutions: 1) Omit the confusing phrase "..., or the specific entry." 2) suggest the obvious: "...or the name of a specific entry". Here's 3) "...or the headword of a dictionary entry or its equivalent". 4) "...or any identifying word or succinct phrase." Lots of options. I don't understand why you are fighting for the existing mess.
Next, to the argument: I disagree on both points.
First, document/article/section/clause is entirely equivalent to a chapter number. I believe most editors will understand that whatever numbering system the source uses should be adopted in citing the source.
"verbatim quote", "cited passage", "number of the table/image", "dictionary definition", "chapter/section number/name", "paragraph number"...Any or all of the above come to mind. Just don't use "entry" alone.FaviFake (talk)12:32, 12 October 2025 (UTC)[reply]
A "dictionary definition" would actually be a "verbatim quote", right? That sounds needlessly complicated when just specifying the entry (i.e. the headword) is sufficient.Gawaon (talk)15:13, 12 October 2025 (UTC)[reply]
I think there is a fundamental misunderstanding of "dictionary entry". I gather that some editors believe that "dictionary entry", or "entry" for short, is theheadword alone. However the word "entry" refers to all of the content that a dictionary gives for the headword. From that same source:
entry Entries are the primary building blocks of the dictionary. Each entry represents all the meanings of a given headword, throughout its recorded history. The entry is structured to show the evolution of meanings and uses over time. In most entries there is also a pronunciation section where relevant, an etymology section, and various other sections. Homographs are treated as separate entries.[2]
Thus "dictionary entry" won't solve this issue. We could use "headword of a dictionary entry", "name of an entry", or other wording that clarifies the intent to identify rather than quote the item.Johnjbarton (talk)15:46, 12 October 2025 (UTC)[reply]
For dictionary or glossary entries, 'Chicago Manual of Style' used to recommends.v. which meanssub verbo orsub voce, Latin for "under the word".[3]. However, §14.130 of the 18th edition abandons that recommendation, and suggests giving the headword of the entry in quotation mark. Some of their examples are
Encyclopaedia Britannica, 15th ed. (1980), under "salvation."
Dictionary of American Biography (1937), "Wadsworth, Jeremiah."
I don't really like "specific entry" because people are obviously confused by that. The passage I cited above from Chicago says "item". I'm not sure that would be understood in this context. I also agree with Blueboar's comment, 'What the f*ck is a “headword”?'Jc3s5h (talk)15:59, 13 October 2025 (UTC)[reply]
If there are no page numbers, whether in ebooks or print materials, then you can use other means of identifying the relevant section of a lengthy work, such as the chapter number, the section title, or the main word at the top of an named entry.
I added the word "named" because the OED source which is quoted above is specific for a dictionary and I think adding "named" clarifies the context for this third item on our list.Johnjbarton (talk)16:24, 13 October 2025 (UTC)[reply]
If there are no page numbers, whether in ebooks or print materials, then you can use other means of identifying the relevant section of a lengthy work, such as the chapter number, the section title, or the main word at the beginning of an named item.
I'd say "headword" wins over "main word at the beginning of a named item" and "brief identifier" for clarity, conciseness, and precision.Gawaon (talk)07:46, 14 October 2025 (UTC)[reply]
Meh… Even with the link, I find “headword” obscure and confusing. However, I can’t think of another word, so I won’t object strongly.Blueboar (talk)15:28, 14 October 2025 (UTC)[reply]
Since many of the things that need to be pointed at don't have headwords, I don't think that's an adequate substitute. "...such as the chapter number, section title, headword, or other brief identifier" would work for me. There is no rule that says we're limited to three examples.WhatamIdoing (talk)18:07, 14 October 2025 (UTC)[reply]
To avoid me getting into a big argument elsewhere, please would someone tell me whether they think this[4] edit is reasonable?
To me, it is entirely pointless to shorten the name of the publisher when the original reference writer either used a template or had the original work in front of them.ThoughtIdRetiredTIR22:12, 11 October 2025 (UTC)[reply]
It's a matter of taste but I think the edits are reasonable. "Penguin" is the part of the publisher name that actually identifies the publisher; "Penguin Books" would also be reasonable; "Penguin Books Ltd." adds information that is not so much part of the name but a description of what type of corporation they are. Editors could in good faith choose to use any one of those three forms. Prior to the edits the names were listed inconsistently ("Ltd." for one publisher, "Limited" for another, just the shortened form "Scribner's" for another not visible in the diff) so it would be difficult to argue that there is already a consistent style choice that should not be changed without discussion. —David Eppstein (talk)22:36, 11 October 2025 (UTC)[reply]
I'd have kept it to Penguin Books myself, but Limited, LCC, Ltd., Private Ltd., GmBH, etc... are just corpolegal gibberish no one cares about.Headbomb {t ·c ·p ·b}22:44, 11 October 2025 (UTC)[reply]
Yeah, just "Penguin" or "Penguin Books" is entirely sufficient, there's no need for more. As I remember it, reliable style guides like the CMOS recommend omitting stuff like "Ltd." in such cases.Gawaon (talk)07:31, 12 October 2025 (UTC)[reply]
IMO the company names are basically obviously superfluous. Books/Press/Publishing should generally be kept as it helps to distinguish and is also generally the most expected name.PARAKANYAA (talk)22:45, 11 October 2025 (UTC)[reply]
What is interesting is to use the isbn to generate the reference with{{citebook}}. Then you discover that the publisher ofthe Bombing War is Allen Lane, an imprint of Penguin Group. So this is a case of impetus to impose a style failing on the accuracy of who the publisher is. The imprint name can often imply (or otherwise) the reliability of a source.
Incidentally, there is a difference between "Ltd" and "Limited" in the UK. This is not stylistic inconsistency, it is a matter of fact. It is always incorrect to swap one for another. Which is why I would resist omitting this element of a name.ThoughtIdRetiredTIR23:02, 11 October 2025 (UTC)[reply]
I'd agree with David Eppstein, this is a matter of taste. I wouldn't suggest going around enforcing one style or the other, but otherwise it's of little importance. --LCUActivelyDisinterested«@» °∆t°23:10, 11 October 2025 (UTC)[reply]
A quick web search suggests that Limited and Ltd are the same in the UK, except that whichever the company uses, they're expected to be consistent about it. That doesn't sound like something we need to worry about, though.WhatamIdoing (talk)04:23, 12 October 2025 (UTC)[reply]
The point is you don't distinguishing between Penguin Books Ltd, Penguin Books Limited, Penguin Books LLC, Penguin Books GmBH, etc. because these are neither meaningful distinctions, nor do they represent different entities than Penguin Books <nothing>.Headbomb {t ·c ·p ·b}05:15, 12 October 2025 (UTC)[reply]
All of those (assuming any of them exist) are the same asPenguin Books as far as we're concerned, but the first two in your list would refer specifically to the British corporation, the third to a US corporation, and the fourth to a German corporation. We don't care, but a contracts lawyer probably would (and shouldn't be using a citation in a Wikipedia article to figure out which entity they're dealing with, so again:we don't care, even if a very small fraction of our readers might,).WhatamIdoing (talk)16:37, 13 October 2025 (UTC)[reply]
I wanted to add a link to the source in the article, but it doesn't add and I'm not sure of its significance. However, in terms of citation, it matches word for word.Smart Andrew (talk)20:33, 12 October 2025 (UTC)[reply]
Our advice here says that weshould use formulations like "Researchers announced the new tissue type in 2012". Our advice at{{who}} says that weshould not make 'attributions to vague "authorities" such as "serious scholars", "historians say", "some researchers", "many scientists", and the like', exactly in contradiction to this advice. AndMOS:WEASEL adds that "A common form of weasel wording is through vague attribution, where a statement is dressed with authority, yet has no substantial basis". The existence of a citation after the sentence is not enough; it will generally support the claim that certain specific people announced the tissue type, but not that a general class of "researchers" did so. The attribution should be made specific, in-text, not indirectly through parsing what the footnotes might mean.
Is there some way of resolving this by choosing better examples here? Or is this a fundamental contradiction where the advice here really is what it seems to be, to change specific attributions into exactly the kind of attribution to vague authorities that MOS:WEASEL argues against? —David Eppstein (talk)21:42, 14 October 2025 (UTC)[reply]
"A 2005 study found that anemones grew and regenerated faster in the presence of clownfish groups, and attributed this to ammonium from clownfish waste."
"The archaeologists who led the excavations at Tell es-Sakan, Pierre de Miroschedji and Moain Sadeq, proposed that there were three areas of Egyptian expansion into the southern Levant during the late 4th millennium BCE, and Tell es-Sakan was one of the major settlements in the region."
"Peter Waddington, in his study of policing, writes that the police's "reputation for restraint received dramatic vindication by the way in which two highly publicised sieges were handled by the Metropolitan Police"."
"The Climate Prediction Center determined four primary factors driving the season's activity: the Atlantic multidecadal oscillation, the reduction of atmospheric convection in the tropical Pacific, record-high sea surface temperatures in the tropical Atlantic and Caribbean, and conducive wind and pressure patterns across the western Caribbean and Gulf of Mexico."
"Among more recent comments on the book, Daniel Hahn in The Oxford Companion to Children's Literature (2015) writes that sentimentality plays a larger part in Through the Looking Glass than in Alice's Adventures in Wonderland."
Researchers announced the new tissue type in 2012.
is a both a weasel andnews. An encyclopedia should include dated items a part of a history which requires context to explain significance (which would of course be available in the secondary reference). All of the examples posted by RjjIII are better.Johnjbarton (talk)02:20, 15 October 2025 (UTC)[reply]
We should always attribute inline for biased statements of opinion asWP:INTEXT says. For other types of statement I think it's up to editor discretion and consensus at the relevant talk pages. If the wording used implies that an opinion is widespread, but in fact that is not the case, then that has to be fixed, and attribution is one way to fix it, but it's not the only way. A relevant recent discussion about an edit to CLOP ishere, and the discussion it refers to at WT:CITE ishere. I think the force of WEASEL should be that omitting attribution is a tactic that can (and should not) be used to inappropriately give authority to some statements, not that omitting attributions is always a bad thing.Mike Christie (talk -contribs -library)02:41, 15 October 2025 (UTC)[reply]
Even for biased statements of opinion, we shouldn't always attribute inline if there are big groups holding the opinion. I often use wording like "arguments in favour are X" and "arguments against are Y", without name dropping like our policy seems to suggest.—Femke 🐦 (talk)16:56, 15 October 2025 (UTC)[reply]
For something that's really an opinion (e.g., "This is a good book", "This politician is important"), then I think the hierarchy ought to be something like this:
Ideal: Talk about established groups associated with known viewpoints.
Republocrats say that this bill solved the fiscal problem, butDemicans opposed it as merely postponing a problem.
Therock opera was praised by critics as a light-hearted romp, but was criticized as irreverent by religious organizations.
Good enough: Identify a couple of individual people and their views.
Alice Expert said that the book was "a valuable contribution to our understanding of just how big the Sun actually is".
Bob Business said that blue-green widgets are anew preppy aesthetic.
Bad idea #1: Hyping the people who are given as examples.
Book expert Prof. I.M. Portant, chair of the Learned Department at Big University, said that it is a good book.
Chris Celebrity, who is famous for having cameo parts in television shows, said that it is a good video.
Bad idea#2: UsingWP:INTEXT to drawWP:UNDUE attention to small facts.
According to a 2016 non-randomized internet-based survey sponsored by Good Organization of people who self-report having autism and who were recruited primarily through social media advertisements, the organization's own membership, and word of mouth, most people with autism like watching videos online.
Oh, and to be clear, the frequency at which something does/doesn't happen is actually a fact instead of an opinion. An opinion would sound like "Wikipedians spend too much time on long arguments" or "Wikipedians have unreasonably long arguments over unimportant things".WhatamIdoing (talk)20:08, 16 October 2025 (UTC)[reply]
Like Johnjbarton I see this form a lot, often unrelated to Wikipedia, in examples that say some position, terminology, or methodology is common or widespread but sourced only to one or more individual publications that take that position and do not say anything about how common it is. —David Eppstein (talk)21:42, 16 October 2025 (UTC)[reply]
Really? Are you sure? Here is the reference from your sandbox:
{{cite news|url=https://www.opb.org/article/2025/10/19/a-peaceful-day-of-no-kings-protests-across-oregon-ends-with-a-show-of-force-in-portland/|url-status=unfit|title=A peaceful day of No Kings protests across Oregon ends with a show of force in Portland|work=[[Oregon Public Broadcasting]]|date=October 19, 2025|access-date=October 26, 2025}}
'requiresurlandarchive-url' – emphasis in original
Your template does not have an archive snapshot so there is nothing for the control-switch parameter|url-status= to switch to. But, if you add|archive-url= and|archive-date= then you get summat that looks like this:
{{cite news|url=https://www.opb.org/article/2025/10/19/a-peaceful-day-of-no-kings-protests-across-oregon-ends-with-a-show-of-force-in-portland/|archive-url=https://web.archive.org/web/20251027231843/https://www.opb.org/article/2025/10/19/a-peaceful-day-of-no-kings-protests-across-oregon-ends-with-a-show-of-force-in-portland/|archive-date=2025-10-27|url-status=unfit|title=A peaceful day of No Kings protests across Oregon ends with a show of force in Portland|work=[[Oregon Public Broadcasting]]|date=October 19, 2025|access-date=October 26, 2025}}
What do folks think aboutthis effort to indicate primary sources with a symbol, and place them in their own separate section? It's like a little ghetto for primary sources. As far as I know, no other editor has used this style. I don't think it serves the reader who will end up questioning whether the primary sources are suitable. In my view, the references should be presented all together, no matter whether they are primary, secondary, tertiary or whatever. The job of questioning whether a primary source is appropriate should be the responsibility of the Wikipedia editor who cites it, not the reader.Binksternet (talk)17:38, 4 November 2025 (UTC)[reply]
I don't understand the point of segregating sources by these definitions and it doesn't seem to be a useful distinction for readers. (Readers who aware of the types of sources can identify them for themselves.)Schazjmd(talk)17:55, 4 November 2025 (UTC)[reply]
I don't think the article even gets the classification of primary or secondary correct. For example, thefirst so-calledprimarysecondary source is basically the author, Gerald Dih, giving his opinion about whether the album is any good. That makes it a primary source.Jc3s5h (talk)19:22, 4 November 2025 (UTC)[reply]
I want to mention that splitting primary and secondary sources in the bibliography is common in articles on classical studies. I would even say that doing so is a best practice.Ifly6 (talk)19:51, 4 November 2025 (UTC)[reply]
OTOH, it probably makes more sense in an academic discipline with a clear and useful distinction between the types, versus Wikipedia where "primary" too often means "I don't like this source so I want to denigrate it in the hope it will be removed" or "I don't like this article, so I'll claim the sources are primary which means it doesn't passWP:N".Anomie⚔01:51, 5 November 2025 (UTC)[reply]
Yea it's also often the case in academic classical studies that the bibliography includesonly modern sources. You're just supposed to know what the citations refer to for the ancient ones.Ifly6 (talk)03:26, 5 November 2025 (UTC)[reply]
I agree the distinction makes indeed a lot of sense when dealing with older historical or literary topics, where "primary" effectively means something like "published more than 300 (or 1000) years ago". It makes less sense when referring to recent topics, where the distinction between primary and secondary is much less sharp.Gawaon (talk)08:38, 5 November 2025 (UTC)[reply]
In any case, regardless of whether or not one groups them in different sections in the bibliography, this marking of primary sources with something like a dagger symbol seems highly unusual and I can't think of any good purpose it could serve. If it's supposed to signal "that's a primary source, don't trust it!" then, of course, it's actually the editors' job to figure out whether a source seems reliable enough to use in the article or not. If yes, then no special "warning" symbol is needed; if no, then it shouldn't be used at all.Gawaon (talk)08:43, 5 November 2025 (UTC)[reply]
I'd just like to add that in the sciences the majority of references you'll see in academic writing are, in fact, primary sources. A secondary source, e.g. a review article, is generally indicated as such. For example, after the title you'll often see "(Review)" added in parentheses. Personally I'm used to seeing mostly primary sources, but I can see the rationale of not emphasizing them here since it's a more general audience.BetsyRogers (talk)00:02, 17 November 2025 (UTC)[reply]
No, that's not how these terms are used, check outWP:PSTS if you didn't yet. Any academic piece of writing must be written with some distance to the subject matter, which makes it secondary. If it's primary (say the autobiography of someone directly involved), then it's not strictly speaking academic writing. Primary literature usually isn't peer-reviewed, while academic (secondary) literature is. As usual, there are borderline cases, but the general distinction is clear enough.Gawaon (talk)08:23, 17 November 2025 (UTC)[reply]
No, that's not how these terms are used You left out "on Wikipedia". We have co-opted these terms and given them somewhat different meanings and implications.Anomie⚔13:25, 17 November 2025 (UTC)[reply]
You may have a point there, especially when it comes to STEM fields where "primary sources" in our sense don't really exist. However, I think that our usage is close enough to the usage established in history, literature, and other humanities, where primary sources are generally non-academic (say a novel about which researchers may write, or letters and diaries from people directly involved in a historical event).Gawaon (talk)15:59, 17 November 2025 (UTC)[reply]
We still have differences, and we don't even agree on them. For example, some consider editorially reviewed news articles about current events as secondary, while others insist they're primary and that secondary sources won't show up until years later. Personally I think we'd do best to demoteWP:PSTS to an essay and have the actual policies and guidelines focus on what really matters like "significant coverage in reliable sources that are independent of the subject" instead of ever mentioning the words "primary" or "secondary". But that's unlikely to ever happen, since too many people like being able to say "it's primary" to mean "I don't like this source for some reason" rather than actually looking into reliability and independence.Anomie⚔17:24, 17 November 2025 (UTC)[reply]
Apparently there are field-specific definitions of a primary source. But using Wikipedia's definition of a primary source (which I checked out a long ago), any published original research is a primary source.
"Primary sources are original materials that are close to an event, and are often accounts written by people who are directly involved."
If someone does the research themselves, analyzes the data, draws conclusions from it, and publishes it, then they are definitely directly involved. I've asked other editors about this to be sure I understood Wikipedia's definition of primary sources, and everyone I've asked so far has concurred that research papers are a primary source.BetsyRogers (talk)01:48, 18 November 2025 (UTC)[reply]
But by that interpretation, wouldn't a historian who researched an event or period also produce "primary sources" because they did the research themselves (visiting archives, studying original sources from those involved or from the time it happened, etc.), then drawing their conclusions and publishing their findings? This would make the distinction between "primary" and "secondary sources" that's usual in history and other humanities completely disappear (merging them both into "primary sources"), so consider me unconvinced.Gawaon (talk)08:17, 18 November 2025 (UTC)[reply]
I'm not trying to convince you of anything, just explaining what constitutes a primary source in scientific fields. I guess if you traveled back in time to directly observe an event or period, then traveled back to the present and published your findings, that would be a primary source. (Alternatively, and with much less effort, you could just Google "what is a primary source in science").BetsyRogers (talk)08:33, 18 November 2025 (UTC)[reply]
UnderWP:CITEVAR, I think it would be difficult to prohibit it (as a general case; getting it removed from a single article would require only an ordinary, consensus-oriented discussion on the article's talk page).
I was not thinking about that and more so grouped them based on independence, I'll combine the sources back together in one list then.Popturtle (talk)09:17, 5 November 2025 (UTC)[reply]
This seems an impossible suggestion to me. Many empirical empirical articles start with a background/literature section making use of that part a secondary source; while it continues presenting new empirical data which makes use of that part a primary source. Thus a single paper can be used in the same wiki article as both a secondary and a primary source. It becomes even more complicated when we look at synthesis from the literature in such articles which can be argued to be either primary (especially if creative combinations are made) or secondary if it is a structured logical addition of what is already known. My head already hurts thinking of all the endless discussion this idea is going to provoke.Arnoutf (talk)18:22, 17 November 2025 (UTC)[reply]
Greetings and felicitations. I thought I recently came across a passage in the MOS or similar document that preservation of a reference's title's original punctuation was not necessary (at least to some degree), but I can't find it. Did I hallucinate it or conflate it with something else? —DocWatson42 (talk)05:44, 12 November 2025 (UTC)[reply]
We are certainly allowed to convert curly quotes to straight quotes. More of a gray area: If a work has a title and a subtitle, are we allowed to standardize the punctuation between them (common choices I've seen include a colon, period, or dash, but sometimes they are just separated by whitespace)? What about (the case I think DW has in mind) changing double quotes to single quotes to avoid nested double quotes? —David Eppstein (talk)06:50, 12 November 2025 (UTC)[reply]
The first and last cases I do automatically (patrolling "quote=" fields for mistakes is a small specialty of mine). For subtitles (in English) I assume that the punctuation is a colon unless the source itself states otherwise. I maintain archaic punctuation (e.g. semicolon em dash) when I come across it. I also convert hyphens to em and en dashes where appropriate (I know that German uses en dashes for ranges, but I don't know enough about French punctuation to make changes), and the same for "x" to multiplication signs. I'm wondering about any other cases (and your thoughts on the preceding types), and if there actually is or was such a statement, or a statement against it. —DocWatson42 (talk)07:10, 12 November 2025 (UTC)[reply]
And if corrections are permitted (as perWikipedia:Quotations#Formatting: "Trivial spelling or typographical errors that do not affect the intended meaning may be silently corrected."). (I'm opposed to correcting the spelling of titles, so that the reference is searchable/reproducible.) —DocWatson42 (talk)07:22, 12 November 2025 (UTC)[reply]
That passage refers to quoted text, it doesn't refer to the titles of cited works and I'd say that it shouldn't apply there. However, changing the type of quotation marks and dashes used within titles is trivial, happens all the time, and is nothing to worry about. (Indeed, changing curly to straight quotation marks is needed per our MOS and changing double to single ones is often needed to ensure proper nesting.)Gawaon (talk)08:17, 12 November 2025 (UTC)[reply]
In the "Citation types" section, under "general references" it says: "Ageneral reference is a citation that supports content, but is not linked to any particular piece of material in the article through an inline citation. General references are usually listed at the end of the article in a References section."
Is this right? I've never heard of putting a "general" reference (i.e. not a source used to verify specific content in the text) in a References section with other citations. Where in the numerical list would it even go? Everything else that I've seen/read indicates that any material that isn't a citation actually linked to the article should go in a separate section (further reading, external links, etc.).
Thanks. (I was editing my question when you replied, just to clarify a bit, but it's still the same question.) But is that still a current recommendation? If not, couldn't it be updated and include a footnote saying basically what you said?BetsyRogers (talk)23:50, 16 November 2025 (UTC)[reply]
See the following sentence about "underdeveloped articles" clarify - it's better than nothing but not typical for higher quality articles. AsWP:GENREF notes, people will generally rework them as an article improves.Nikkimaria (talk)04:42, 17 November 2025 (UTC)[reply]
how to add this insert to my draft (for example, the king of England from such to such a year is the predecessor/ successor) if something is not the king but the baron, I make an example of the king to make it clearer what I'm talking aboutSmart Andrew (talk)07:58, 27 November 2025 (UTC)[reply]
The redirectExcessive citations has been listed atredirects for discussion to determine whether its use and function meets theredirect guidelines. Readers of this page are welcome to comment on this redirect atWikipedia:Redirects for discussion/Log/2025 September 22 § Excessive citations until a consensus is reached.Thepharoah17 (talk)07:12, 22 September 2025 (UTC)[reply].png)
N In an article published inThe Lancet in 2012, researchers announced the discovery of the new tissue type.[3]
Y Researchers announced the new tissue type in 2012.[3]