Wikipedia talk:Citing sources/Archive 48

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

Since the |first= and |last= parameter values render as "last, first" no matter the sequence they occur in the code, it makes sense to code them in the sequence seen in the source. You don't have to mentally rearrange the name, making it both easier and in some cases less error-prone.

And yet most editors code this as |last=|first=. I can only surmise that some do it per the coding examples, others because they think that's required to get the "last, first" rendering, and others because it's safer to follow the crowd.

Considering sources of all types, it's my belief that a large majority show authors' names as "first last"; certainly all news sources and most book sources do. Therefore I propose changing all examples in template doc and other pages to show |first=|last=. I would do what I can to help with that. ―Mandruss ☎ 16:32, 14 August 2019 (UTC)

That sounds like a make-work-for-the-sake-of-making-work kind of project. As you noted, the order of |firstn= and |lastn= in the code does not matter with regard to the rendering. I think that the order in the code is only important when cs1|2 templates are used in bibliographic listings to make it easier to get the alpha order right. In bibliographic listings, |lastn= should be the first parameter in the template:

{{cite ... |last=Surname |first=Given Name |...

—Trappist the monk (talk) 16:39, 14 August 2019 (UTC)

As I indicated, it makes a difference as to ease-of-use, even if not a huge difference. If I have the necessary access, I'll do all the work myself; perhaps I'm permitted to "make work" for myself? ―Mandruss ☎ 16:42, 14 August 2019 (UTC)

I disagree with the concept of coding the first name first. Even if an article does not currently have a bibliography in alphabetical order, it may in the future, as the article gets larger or is more carefully edited. We do not have any tool I know of to automatically put a bibliography in alphabetical order; it's done manually. If the code for the citation begins with the last and first name of the authors, as shown in the publication, followed by the date, it will be easier to alphabetize the bibliography.

In terms of making things easier for the reader, if the reader needs to look up the work in an electronic card catalog or the like, the reader will probably be prompted to enter the last name followed by the first name, so this makes it easier for the reader too. By the time the reader gets the physical work in hand (or on screen) all the lookup work is finished. Jc3s5h (talk) 16:46, 14 August 2019 (UTC)

it will be easier to alphabetize the bibliography. And the ease-of-use for that very rare case outweighs the ease-of-use for everyday work by thousands of editors?

In terms of making things easier for the reader - Sorry, I don't follow. Readers don't see the code, this is only about the code. ―Mandruss ☎ 16:54, 14 August 2019 (UTC)

(edit conflict) Agree that it makes it much easier for editors if the first field is |last=, and it makes no difference to readers. It is the normal order for all manual filing systems and for ordering books on shelves. Change would be disruptive and counter-productive. Martin of Sheffield (talk) 16:55, 14 August 2019 (UTC)

I agree with Mandruss in both of his/her edits above ~ most sources I work with state the authors name as |first= |last= |, mentally you have to switch the names. It's much easier on your brain or lack of ~ ~mitch~ (talk) 17:05, 14 August 2019 (UTC)

~mitch~ is right if the editor is entering the information directly from the source. But that involves flipping between various pages, either physically, or onscreen. The editor may have obtained access to the source from some online catalog, which conveniently has almost all the information you need for the citation in one page. Or, the source may provide suggestions on how to cite it, which also are all in one spot. The catalog or suggested citation will probably list the last name first. Jc3s5h (talk) 17:20, 14 August 2019 (UTC)

Again, you're putting outlying cases before far more common work. ―Mandruss ☎ 17:34, 14 August 2019 (UTC)

The software I sometimes use for generating citation templates puts them in last-first order. As already stated above, I find this order helpful when alphabetizing citations. But when I create templates by hand (usually by copying and pasting a suggested citation format from some source and then adding the parameter names around the copied and pasted text) I leave them in the order I found them. I I don't think we should be encoding, stating, or enforcing any preference for parameter ordering. They have to be ordered somehow in the documentation, but there should be no suggestion that the documentation order is in any way significant. —David Eppstein (talk) 17:33, 14 August 2019 (UTC)

there should be no suggestion that the documentation order is in any way significant. I totally agree. No more suggestion than already exists. ―Mandruss ☎ 17:35, 14 August 2019 (UTC)

I agree with David: we should NOT be enforcing any kind of parameter order. (And the bot drivers should not mess with such!) There are some aspects of ordering that might be preferable in some cases, but we should NOT be enforcing any kind of One and Only True Style.

As to preferences: I have strong preferences on some aspects of parameter ordering (e.g., having last= and first= on the same line), but last-first order? It depends, and I swing both ways. I scrape off a lot of author listings from journals, and I generally go with the order found. For me the most important point is to have the last names aligned vertically to more readily check them, for which last-first works. But I would prefer to have the first names (or initials) aligned as well, which, in the presence of a long last name, pushes the first names over to the right. Alternately, if (and for me, the general case) initials are being used, then first-last is reasonable. But I usually follow what is given. As to what cases are "more common": your mileage undoubtedly varies. Perhaps what the examples should say is that the order doesn't matter, and can be done as may be most convenient. ♦ J. Johnson (JJ) (talk) 20:49, 14 August 2019 (UTC)

Nobody has proposed "enforcing" anything here, least of all me. Please read the opening comment, in which I clearly articulated what I propose. ―Mandruss ☎ 21:09, 14 August 2019 (UTC)

Err, "Therefore I propose changing ... and other pages to show |first=|last=" – your words. The examples are less significant, what would be unacceptable would be steamrolling down bibliographies which are in alphabetic order of last names and changing them. You'd just be making maintenance of references harder for no good reason. Martin of Sheffield (talk) 21:28, 14 August 2019 (UTC)

[This section was started as a subsection of the previous discussion (#Citation extremeness). But that discussion has gotten so far out in the tules that this discussion should be allowed a fresh start, so I am boosting it up a level. ♦ J. Johnson (JJ) (talk)]

I just made an edit that shows the potential limitations of bundling references in a citation. In two citations, most of the references given are the same. But not entirely, and the order given is different. Thus, simply re-using the citation doesn't work. Fully giving the references each time does work, but not typical convention. The only choice then to avoid several citations is to manually type out the sources - this isn't so helpful if the reader wants to jump to the full version of that source via an internal link.--3family6 (Talk to me | See what I have done) 03:46, 20 August 2019 (UTC)--3family6 (Talk to me | See what I have done) 03:46, 20 August 2019 (UTC)

That's not the only choice. You can have two separate section, one for footnotes (with short references in

WP:HARV style) and a second one for the full references to which the footnotes refer. In that style, you would only spell out the details of each reference once (in the full reference page), and you can choose within each footnote what order to cite the references. This style choice is independent of whether you bundle the references or choose to use separate foornotes for each short reference. —David Eppstein (talk

) 23:17, 20 August 2019 (UTC)

Okay, that's what I've done before. I'm realizing now how you can included multiple references inside a particular short foot-note.--3family6 (Talk to me | See what I have done) 23:38, 20 August 2019 (UTC)

Your comments are nonsensical. For instance: "references" often refers to "citations", or notes, or about six other things, so it is not all clear what you are referring to. As discussion is futile unless the terms we use have the same meaninging for each of us. I suggest we should take care to use the key terms as follows.

"References" is too commonly used in too many different ways to have any reliable sense, and therefore should be avoided. "Citation" generally refers to a full citation (what the Chicago Manual of Style calls a "full reference") that includes the full bibliographic details of a source. On WP these are most often created using the {{

harv

}} family of templates. If the full citation is generated using {citation}, or using a {cite} template with the |ref= parameter, the short-cite generated by a properly done {harv} template will automatically link to the full citation.

Here is a key part: the notes (a.k.a. "footnotes", "end notes") created using <ref> tags are NOT citations! They are just a place (again, automatically created) where you can put a citation, or comments, or just about anything except another note. (A fair interpretation of your "bundling references in a citation" is that you mean putting <ref>...</ref> tags inside a {citation} template. Which is such a crazy idea I just had to try it. Nope, tagged as an error.)

One more thing: it is not a merely "typical" convention, but a universal practice that (in a given unit) there is only one full citation for each source.

What you ran into is called the "re-use" problem, where you want to "reference" (cite) a source from more than one location. There is a very simple solution: short-cites. Use them as often and where ever needed, they will link to the indicated full citation whether it is in a note somewhere, or (what David is talking about) collected with other full citations in their own section.

Strictly speaking, what we are bundling in this manner are citations. To bundle references (where notes are being referred to) is indeed very limited, and best not attempted. The short, simple answer here is: bundle citations (full or short), not notes. ♦ J. Johnson (JJ) (talk) 19:54, 21 August 2019 (UTC)

Woah, steady on! Forgive that my terminology is muddled, as you said, the terminology as practiced by Wikipedians is muddled. I provided the edit to make clear what I was trying to do. I appreciate the clarification, believe me I'm not intentionally trying to be muddled, and this is why I provided a link to the edit. The issue was explained to my satisfaction by David Eppstein and the documentation at the pages for the short-footnote and Harvard reference templates.--3family6 (Talk to me | See what I have done) 20:23, 21 August 2019 (UTC)

I don't believe anyone is trying to be muddled. But there is so much muddledness practiced on this topic that it can't be avoided. Supplying the link was good, because then I could see what you were trying to describe. And as I said in my edit summary: the solution is simple – once the terminology is sorted out. ♦ J. Johnson (JJ) (talk) 21:41, 21 August 2019 (UTC)

More than three citations destroys the readability for readers. Guidelines should not allow editors to add 5 or more citations after each sentence. One citation is usually plenty. More than 10 citations for one claim is craziness. QuackGuru (talk) 18:15, 14 August 2019 (UTC)

Three is my rule-of-thumb limit, but not without allowance for warranted exceptions. We have essay Wikipedia:Citation overkill, which you are free to point to in discussions, but we will never have a guideline on this. ―Mandruss ☎ 18:19, 14 August 2019 (UTC)

And we have the

citation bundling technique, which allows unlimited citations to be combined into a single citation number. ―Mandruss ☎

18:21, 14 August 2019 (UTC)

Essays don't stick. Let's start with 5 or 10 is the max for this guideline. It is getting to the point of disruption. QuackGuru (talk) 18:23, 14 August 2019 (UTC)

@

citation bundling technique if you think there are too many cites for one sentence. ~mitch~ (talk

) 19:26, 14 August 2019 (UTC)

But the number of citations is actually not the problem. QuakGuru is most likely referring to the note links – the "^[1][2][3]" numbers that link to the notes ("footnotes", "end notes"). Which are NOT citations; they are just where most editors dump their citations. As it is, I think it would be preferable to have only one note – and therefore only one note link – at any point, with all relevant citations, comments, figures, etc., contained in that single note. Of course, this is quite impractical the way most editors do citation. So what really should be fixed is: the way most editors do citation.♦ J. Johnson (JJ) (talk) 21:19, 14 August 2019 (UTC)

• Johnbod's Law says "5 refs on a line is almost always a sure sign of trouble". This usually occurs in areas of controversy, where statements are contested, typically in medical areas or controversial assertions of nationality (especially in the Balkans). QuackGuru's own editing style may have provoked some of these build-ups, as he tends to argue about references. I certainly approve of bundling, but it isn't as practicable for medical articles, given the prevailing style of medical referencing. Johnbod (talk) 21:28, 14 August 2019 (UTC)
  • The build-ups are three citations per claim for controversial claims, for me. Bundling two citations is the max, for me. QuackGuru (talk) 21:37, 14 August 2019 (UTC)

One citation per claim is the general rule I follow. I noticed others like multiple citations or whatever after a sentence. I noticed most of the additional citations do not even verify the claim. It is tantamount to ref spam. If more than one citation is required for a sentence then just place the citation where is verifies the claim: "The cytotoxicity of e-liquids varies,[27] and contamination with various chemicals have been detected in the liquid.[28]" QuackGuru (talk) 21:37, 14 August 2019 (UTC)

Absolutely, and we already have a guideline somewhere to that effect. Wouldn't your time be better spent fixing those problems, with edit summary links to that guideline for the education of other editors? ―Mandruss ☎ 21:40, 14 August 2019 (UTC)

You know @QuackGuru: instead of just copying the text from article with the ref # in the brackets ~ it would help if you actually copied the edit with the source ~ I might be able to understand better what you are talking about ~ example ~ I think this is what you are saying ~ correct me if I am wrong ~

• Austin–Bergstrom International Airport is currently in the final stages of preparing their 2040 master plan. As ABIA continues to experience rapid growth, {...} existing Garage 1 is.^[1][2]

Should have been;

• Austin–Bergstrom International Airport is currently in the final stages of preparing their 2040 master plan.^[3] As ABIA continues to experience "rapid growth",^[4] {...} existing Garage 1 is.

Is that what you are talking about? ~mitch~ (talk) 22:05, 14 August 2019 (UTC)

Yes. Also there are times when too many citations are used at the end of the sentence. Let's limit it to 5, would be a start. QuackGuru (talk) 22:16, 14 August 2019 (UTC)

I'm not in favor of limiting how many sources are needed to correctly cite an article ~ we have other ways, as mentioned above, to handle what you call to many ~mitch~ (talk) 22:28, 14 August 2019 (UTC)

I'm also not in favor of limiting how many sources are needed to correctly cite a sentence. That's not the issue. When one citation verifies the claim adding 5 or more is excessive. Even more than 3 is kind of too many. QuackGuru (talk) 22:35, 14 August 2019 (UTC)

Citations are used for multiple purposes. "Verifying the claim" is only one of them. Another, for instance, is assigning proper credit to discoverers. If I want to say for instance, that after multiple improvements to some bound in mathematics, the best proven bound is something, then I may well want to cite a large number of sources for those multiple improvements. And when I'm listing books authored by the subject of an article, I am likely to cite as many reliably-published reviews of those books as I can find, rather than just a single review, because I am documenting the amount of critical attention the book has received rather than verifying the existence of the book (even a single review would be overkill for that verification task). Of course, in these cases I would probably do so in a single footnote with some form of citation bundling rather than in multiple footnotes, but you seem to be objecting even to using multiple sources. —David Eppstein (talk) 23:28, 14 August 2019 (UTC)

I never seen a case where 5 citations were needed for each claim. One sentence could need about 20 citations to verify all the claims, but each specific claim would still have only one citation. A single citation for each claim is usually plenty. Never more than 3 for each claim is ever needed. QuackGuru (talk) 02:52, 15 August 2019 (UTC)

The first note at Earthquake prediction has five citations. That is because there is some spread in the conception of the topic, and/or the expression of that conception, so some explanation is provided for the interested readers, and for any editors prone to quibble the fine points. But note (ha ha): five citations, but only one note. Now hearken back to QG's opening statement here: More than three citations destroys the readability for readers. Huh? How do citations affect (destroy??) readability if they are in the note? QG: aren't you really complaining about sequences of numbered note-links in the text? That is a different issue than how many sources are need to support a point. These issues are entangled only because of the failure to distinguish citations from notes. ♦ J. Johnson (JJ) (talk) 18:57, 15 August 2019 (UTC)

This was not specifically about "notes". Although if additional citations in the notes are not needed then there can be a set limit. QuackGuru (talk) 20:12, 15 August 2019 (UTC)

Indeed, the number of citations should be based on what is needed, or at least useful. But NOT a "set limit" that has no consideration for need or usefulness. ♦ J. Johnson (JJ) (talk) 21:45, 15 August 2019 (UTC)

Help us understand the real-world situations you're encountering that prompted this request. Are you running into issues with editors who are trying to cite more than three, even after you object? I assume when you see more than three citations, you're reducing them to 2 or 3 and moving on without any issues in most cases. Just trying to understand the real impact here. --GoneIn60 (talk) 03:29, 15 August 2019 (UTC)

This is a long-term problem. Without a guideline limiting the number of citations after a claim things like this happen. More than 5 citations after a claim can be tagged by a bot. First, we need to come to some sort of consensus for the wording in this guideline. QuackGuru (talk) 20:12, 15 August 2019 (UTC)

The example you point to is ludicrous, and such cases should be pulled down on that basis, not because they exceed some arbitrary limit. In that case, also on the basis of OR: it appears someone was trying to establish various claims of "Alvin Duskin is X" by showing multiple instances of "X", instead of showing a source making that claim. Various other objections could be made, and the lack of a "a guideline limiting the number of citations" is NOT the problem. Nor is it a valid solution. ♦ J. Johnson (JJ) (talk) 21:49, 15 August 2019 (UTC)

As an aside, I notice that the article is now both vastly longer and largely uncited.Nigel Ish (talk) 22:10, 15 August 2019 (UTC)

Citing a lot of sources at the end of a new article, during the first hour after creating it, doesn't mean "These citations support the last sentence in this article". It means "Attention, New page patrollers: This guy really is notable". WhatamIdoing (talk) 23:47, 15 August 2019 (UTC)

I've seen overcitation happen in three different types of situations:

• We have information that changes over time - for example, sales figures for a video game. I generally append the latest source to the end of the existing references to update the numbers, in case the others are used, but once that gets past 3-4 refs, then yes one should determine if the earlier references are needed or not (for ref-reuse) and eliminate the unnecessary ones.
• There is a point of contention about a certain statement (like a label) which is the DUE point of view but because some editors keep wanting to insert the FRINGE side, other editors keep bolstering up references to say "Uh, this is not a point to be challenged because of X sourcing having it." Excessive references should not be used to fight that type of battle: editors should select the 2-3 best sources that make the point, or if that can't be determined, bundle-cite them.
• There is a point of contention between two or more valid points or sides , neither failing FRINGE. Now you got two sets of editors doing that. In this type of case, this should mean you are going to have a section in the article that should go into difference between those two points, those references spread among them, rather than trying to bundle them up on a simple statement.
• I completely agree that there should be a limit of something like 3 consecutive in-line refs (inculding any footnotes), and if you go over that, and can't resolve the refs any other way, then bundle-cite them to cut down the size. --Masem (t) 23:08, 15 August 2019 (UTC)

I've seen them too, but are these situations where you advise against it, reduce the number of citations, and then proceed to receive pushback from the same editor? Quack brought this up under the premise that it reduces readability, but you can simply bundle citations into a single footnote as demonstrated at

WP:CITEBUNDLE. Because this is a viable option, any claimed loss of readability can easily be remedied. What is the expectation here? If we add a limitation, will that really curb the behavior, or in actuality will it be limiting editors in rare situations when they might have reason to provide more than 3 citations? The reason for the push and the expected outcome from a change still isn't clear to me. --GoneIn60 (talk

) 23:28, 15 August 2019 (UTC)

I'm more concerned about the opposite scenario, in which you name five reliable sources, someone argues that a single citation is sufficient and the rest are mere clutter, and then the next person says that the information should be removed because the single source that "consensus" allowed you to retain doesn't prove that the content is

WP:DUE or can't prove "multiple improvements" (to pick up User:David Eppstein

's example).

I suspect that the more common scenario is a straightforward content dispute, in which an editor addresses the "But that can't be verifiable (since it's not what I personally believe)!" comments by dumping a large number of citations on a claim until the others have to admit that the claim is indeed verifiable. And even these don't happen often; the example given is more than a year old.

But as you note, if these disputes are getting solved in practice, then there is no need for a

WP:CREEPy rule specifying the "correct" number of sources, even if we all agreed upon a single "correct" maximum number, which we don't. For example, in the example given, Quack simply blanked 80% of the citations

.

(Bundling is a matter of CITEVAR. If editors don't want to use it, then we won't make them. Besides, bundling is a problem when several of the sources you're citing are already cited elsewhere in the article. It's cleaner to cite them all separately than to make ever-varying combinations in bundles.) WhatamIdoing (talk) 23:58, 15 August 2019 (UTC)

See "One is usually enough, and more than three is usually a sign of problems."

Disputes are not getting solved in practice, so then there is a need to specify the "correct" number of inline citations (including any footnotes). QuackGuru (talk) 14:27, 16 August 2019 (UTC)

That new editor hasn't been on wiki since that advice was posted just yesterday. Someone's "failure" to edit as often as you do is not proof of the "dispute" (it hardly rises to that level) not being resolved. WhatamIdoing (talk) 15:41, 16 August 2019 (UTC)

We do need to keep in mind that there are some statements that I would routine expect to see 2-3 sources used to back a claim, this being when there is some contentious statement, such as "Some consider John Q. Smith to be a white supremacist." - the "some" implicitly means "more than one RS" and that better have 2-3 sources. This is a case where I have seen editors, which wish to contradict that statement, insist more sources be added, editors wanting to keep it add more, and suddenly you have 7-8 references on a point like that. It all depends on context, but in a case like my example, you go through and pick out the top 2-3 RSes that support that and leave it at that. And then drop the rest in the talk page so if it is challenged again, you can point exactly where that was discussed. But this would be helped by having guidance on the max number of cites in a row to avoid citation overload here. --Masem (t) 14:35, 16 August 2019 (UTC)

GoneIn60, simply bundle citations into a single footnote is moving the problem from one place to another. QuackGuru (talk) 14:27, 16 August 2019 (UTC)

But that would fix the "destroys the readability" problem with which you started this discussion.

As to "too many citations": if there are reasons for having, or even not having, five or six or eight citations, then any challenges should be decided on the basis of those reasons, not some arbitrarily set number. If we set a limit (or even a "guideline") then, for some editors, "N>3" will be a brightline basis removing citations, and a new realm for contention. ♦ J. Johnson (JJ) (talk) 21:06, 17 August 2019 (UTC)

Cluttering the reference section by bundling the citations will confuse the reader when they want to click on a source. Bundling multiple sources together into one citation is confusing. It may look like one citation but it contains multiple URL links. That *is* confusing. Most editors don't bundle anyhow.

It is an old realm of contention for providing more than three inline citations for a given claim. So far no reason based on verifiability has been presenting for having more than three inline citations. QuackGuru (talk) 21:34, 17 August 2019 (UTC)

I find nothing confusing about #705 here (permalink). The sources are clearly delineated by bullets. That's an excellent example of a case where more than three (nine, in that case) cites are warranted. It's a highly controversial statement that needs rock-solid sourcing, and that certainly outweighs any unsubstantiated claims about reader confusion. ―Mandruss ☎ 21:58, 17 August 2019 (UTC)

That's a better way to bundle by using bullets.

I was not referring to highly controversial statements. That's not what this is about. For non-controversial claims more than three inline citations has not been shown to be a benefit for readers. QuackGuru (talk) 22:06, 17 August 2019 (UTC)

Wikipedia (en-wiki, at least) hates bright lines, preferring to leave virtually everything to discussion and consensus. Many better cases for bright lines than this have failed. Good luck breaking that pattern. ―Mandruss ☎ 22:11, 17 August 2019 (UTC)

On the other hand, I did point out a non controversial instance of five citations. Quack, are you paying full attention, or just skimming across this stuff? ♦ J. Johnson (JJ) (talk) 00:06, 18 August 2019 (UTC)

I counted only one "inline" citation. See Earthquake_prediction#cite_ref-1. QuackGuru (talk) 00:10, 18 August 2019 (UTC)

That's one note "in-line", containing five citations. Your prior comment (above, at 14:27) is that to "bundle citations into a single footnote is moving the problem from one place to another." So which horse do want to ride: an alleged problem with multiple notes? Or an alleged problem with multiple citations? Do keep in mind that a note (footnote) is NOT a citation (of any kind). ♦ J. Johnson (JJ) (talk) 22:37, 18 August 2019 (UTC)

A set limit of three ([1][2][3]) for unbundled inline citations (not the bundled citations into a single footnote). QuackGuru (talk) 00:07, 19 August 2019 (UTC)

Those are links to three notes. The example I provided is a single note (^[1]). Please do not say "citations" when you mean "notes" (or the links to notes).

Again: citations ≠ notes. ♦ J. Johnson (JJ) (talk) 00:34, 20 August 2019 (UTC)

(To Mandruss's comment) This is probably beyond the specific scope here, but I feel there's something odd about having to have 9 citations in a bundled cite to support a single sentence that contains a known contentious claim. I mean, this is where a statement should be explained if you have that many sources. Not that you don't have that in the sub-article that is linked where you have clearly more space to expand (I don't know if all those cites are used but I would expect them to). this would be the case in the summary article that I would trim those sources to the two-or-three most firm /highest-reliable sources and let the sub article speak to more detail. I am not doubting the statement itself or the sources, just that when you put 9 sources on one short sentence, that feels like POV pushing. Piling up sources without expanding on those sources is sorta avoiding explaining the point. This goes back to one of the cases I explained above. But again, this edges into NPOV to some degree, as well as beyond the issue of "lists of inline cites", so may need a different discussion. --Masem (t) 00:10, 19 August 2019 (UTC)

The prevailing mind-set at that article (my understanding of it, that is) is that few readers will click through to any of the sub-articles, so their existence can't figure into decisions about the root article; i.e. the root article must stand on its own in isolation. And when you devote space to explaining the point, you get pushback per WEIGHT. I doubt there will be any new written guidance in this area, but I welcome you to discuss this at Talk:Donald Trump. ―Mandruss ☎ 18:32, 19 August 2019 (UTC)

See "The statements have been documented by fact-checkers; academics and the media have widely described the phenomenon as unprecedented in American politics.[705][706][373]" The content fails verification. It is a SYN violation to claim "...widely" described...". This requires a rewrite. See Earthquake_prediction#cite_ref-1. Bundling the citations makes it vey difficult to verify which citation verifies which claim. Therefore, it is like a policy violation when a reader can't verify the claim. See "Some consider..." is also a SYN violation when multiple sources are used to verify the word "some" when no single source verifies the claim. QuackGuru (talk) 19:36, 19 August 2019 (UTC)

That is the result of extensive discussion and has consensus. If you wish you can raise objection on the article's talk page. We're off topic here, so that's all I have to say. ―Mandruss ☎ 19:51, 19 August 2019 (UTC)

On the specific point for the Trump article, I agree this is off topic. But there is still something about a bundled site supporting one sentence that has a controversy claim with more than X citation that feels off, but I can't put my finger on it or how to deal with it, but it does come to the issue alluded by QuackGuru, that is like telling a reader "Do you want to verify this fact? That's a needle in this giant haystack over here." WP:V does require giving the reader help, but throwing them N+ citations without clarity is an issue to that end. --Masem (t) 20:32, 19 August 2019 (UTC)

It is near impossible to deal with it unless the WMF hired super admins to enforce the rules. The best I can hope for would be a general statement for limiting the number of sources for each claim. Wikipedia seeks consensus over rightness. QuackGuru (talk) 20:58, 19 August 2019 (UTC)

You stated "That's an excellent example of a case where more than three (nine, in that case) cites are warranted." I disagree that nine citations are warranted for a single claim, especially when the content has failed verification. Those sources and quotes do not verify the entire claim. I don't think a single editor can back-up the claim, with including the unsupported weasel word. Words such as "widely" and "some" should be supported by a single source rather than adding up sources to come to a new conclusion not found in any single source. To summarise my point, I think if we included a general statement for discouraging more than three citations for a single claim would be helpful for our readers. Trying to find a needle in a haystack is the wrong approach. QuackGuru (talk) 20:50, 19 August 2019 (UTC)

All citations should be specific enough to verify the claim, regardless of whether there are other citations nearby that also do. So your (Masem's) analogy of a needle in a giant haystack is off. It's more like finding a needle in a giant stack of needles. —David Eppstein (talk) 20:58, 19 August 2019 (UTC)

There are no citations that verify that specific claim with including the weasel word. But editors seem to like adding up nine citations to come to a new conclusion. The end result. The readers won't find a needle in a giant stack of needles. This brings me back to my point. There is no reason for using so many citations for a single claim based on policy. QuackGuru (talk) 21:07, 19 August 2019 (UTC)

That's yet a wholly third issue related to the matter (using many refs to support an claim like "Many people think...") which, to me, still needs to be dealt with in terms of how many citations to throw into a bundle to try to justify that point, but that I think is also off topic here. There's some allowances for this approach to sourcing such claims, but it should not be "I have a gazillion sources for this, my claim wins." logic that sometimes exists out there. But again, beyond scope of "too many inline citation tags in a row" this discussion started as. --Masem (t) 00:25, 20 August 2019 (UTC)

TL;DR: Bundle citations into a solitary note, or a couple notes, if you need multiple cites. Solution provided.--3family6 (Talk to me | See what I have done) 20:26, 21 August 2019 (UTC)

Exactly. Nothing extreme about it. ♦ J. Johnson (JJ) (talk) 21:43, 21 August 2019 (UTC)

Exactly the opposite. Bundling citations has its own set of problems. Editors should avoid bundling citations when there is a problem with too many citations. For example, see Earthquake_prediction#cite_note-1. That is a good example where bundling causes a verification problem. Nobody knows which citation verifies which claim. Same goes for the 9 citation bundle. A good guideline is when editors don't like it, but they know it is the right thing to do. I think we can work on getting the wording right for a RfC.

See "Editors should be cautious to avoid bundling too many citations together into a single footnote when different citations verify different parts of a sentence."

See "Overcitation happens when more than three consecutive inline citations or notes are used to cite a non-controversial claim." QuackGuru (talk) 22:08, 21 August 2019 (UTC)

Quack: your assertion of a verification problem is bullshit. In the case at hand, all of the citations in note 1 apply to exactly ONE part of the sentence. Your complaints of multiple claims and "different parts of a sentence" are simply inapplicable. Since you are too muddled to understand this I am not going to waste any more time on you. ♦ J. Johnson (JJ) (talk) 23:06, 21 August 2019 (UTC)

QuackGuru has a very valid point. Per

WP:V, we want to provide a citation with as reasonable a narrowness of range of the original source material for the reader to find and verify the material. So if I was using a 1000 page book as a source, I should at least narrow that down to chapter, or if possible, by page, to make it easy to verify. Just giving the book and no further "pointers" to where in that 1000pg book is not useful as a reference. That issue extends if you use more than 2-3 references in a bundled cite, if the references broadly support the sentences they come after. It means the reader has to read all three works to figure out what was supported and where in the references. Sometimes, this can be fixed by unbundling, eg: A statement like "Presidents Carter, Reagan, Bush, and Clinton all has policies related to budget control." where there are four cites for each name, would be better to put each cite after the name, rather than bundling the cites. --Masem (t

) 23:11, 21 August 2019 (UTC)

See "Earthquake prediction is a branch of the science of seismology concerned with the specification of the time, location, and magnitude of future earthquakes within stated limits,,[1]" See "In the case at hand, all of the citations in note 1 apply to exactly ONE part of the sentence." What about the other parts of the sentence? Which citation verifies which claim? The bundling is confusing. It is like a jigsaw puzzle. I am working on a draft for a new article. I only used one citation per claim and for PDF files I provided the exact page number. There are sentences that use more than one citation and for those each citation is placed where it verifies each claim. I recommend a moratorium on citation bundling after observing what is happening. QuackGuru (talk) 23:32, 21 August 2019 (UTC)

I agree that citations for different claims, particularly in different parts of a sentence, should not be globbed together without any indication of where they apply. But neither should should they be strung out in a sequence of notes all at the end of a sentence. In short, that's not a bundling issue, that is Quack jumping from one ill-conceived complaint to another. I do dispute his assertion that note 1 "is a good example where bundling causes a verification problem."; it is entirely invalid. I rather doubt that he has made any valid points in this discussion. ♦ J. Johnson (JJ) (talk) 23:45, 21 August 2019 (UTC)

Just a little help QuackGuru ~ so everybody knows what you are talking about ~ ~mitch~ (talk) 23:51, 21 August 2019 (UTC)

You can also include explanatory text within the note, too. Take a look at these examples: note 79 and note a; note b. If it is ungainly to include the additional explanatory text within the article body, you can include it in the note. "X statement is made in source Y, Z statement is made by page 561 of source A."--3family6 (Talk to me | See what I have done) 00:36, 22 August 2019 (UTC)

I'll copy the examples to here to make it easier to see what I'm talking about:

"The art on albums by Viking metal artists frequently depicts Viking Age archeological finds: Thor's hammers are especially common, but other artifacts such as Oseberg posts and even the Sutton Hoo helmet have appeared.^[1][a]"

"Other European languages, such as

Gaulish are sometimes used.^[b] Heavy metal fans around the world sometimes learn languages such as Norwegian or Finnish in order to understand the lyrics of their favorite bands and improve their appreciation of the music.^[11]

"

References

1. ^ Trafford & Pluskowski 2007, p. 65. sfn error: no target: CITEREFTraffordPluskowski2007 (help)
2. ^ Carver 1998, p. 164. sfn error: no target: CITEREFCarver1998 (help)
3. ^ Campbell 2009. sfn error: no target: CITEREFCampbell2009 (help)
4. ^ Bowar 2011 harvnb error: no target: CITEREFBowar2011 (help); S., Mike harvnb error: no target: CITEREFS.,_Mike (help)
5. ^ Ponton 2010. sfn error: no target: CITEREFPonton2010 (help)
6. ^ Ashby & Schofield 2015, p. 502 harvnb error: no target: CITEREFAshbySchofield2015 (help); Zed 2012 harvnb error: no target: CITEREFZed2012 (help)
7. ^ Ulrika 2014 harvnb error: no target: CITEREFUlrika2014 (help); Slechtvalk 2000 harvnb error: no target: CITEREFSlechtvalk2000 (help)
8. ^ Metal Marc et al. 2002. sfn error: no target: CITEREFMetal_MarcFlex187xRTxHeidendoder2002 (help)
9. ^ Angelique 2005. sfn error: no target: CITEREFAngelique2005 (help)
10. ^ Mulch 2014 harvnb error: no target: CITEREFMulch2014 (help); Weinstein 2014, pp. 66–67 harvnb error: no target: CITEREFWeinstein2014 (help)
11. ^ Rossi & Jervell 2013. sfn error: no target: CITEREFRossiJervell2013 (help)

--3family6 (Talk to me | See what I have done) 00:47, 22 August 2019 (UTC)

Altering citation titles to comply with WP MOS

Hi, what's the consensus on altering original citation titles (e.g. capitalization, punctuation, dashes, or quotation marks) to match Wikipedia MOS or general English style (even if the source is not English)? Say, for example bot edit here where Lithuanian quotation marks „ “ were changed to English quotation marks. I have seen plenty of similar edits. Was there a previous discussion somewhere on this? Renata (talk) 02:19, 27 August 2019 (UTC)

Take a look at

MOS:TITLECONFORM, which support that particular edit. Nikkimaria (talk

) 02:39, 27 August 2019 (UTC)

This guideline, "

Citing sources

", allows the use of any consistent style. If a style from some comprehensive style guide, such as the Chicago Manual of Style is being used, the recommendations for treatment of titles from that manual could be used for citations. Such manuals will probably call for changing foreign titles to be more manageable in an English-language context, such as using quote marks and transliterating non-Roman alphabets.

If the style has been made up for that one article, that's allowed. If no foreign-language sources have been added so far, and you want to add the first one, there is no guidance; do what you want. Jc3s5h (talk) 14:10, 27 August 2019 (UTC)

RfC appears to contradict MOS here

An RfC here, about a citation template and not about MOS, concluded that "names of websites in citations/references should be italicized, generally in line with current practices," which is neither policy nor guidelines and appears to contradict

WP:CITESTYLE

.

WP:CITESTYLE states, "Wikipedia does not have a single house style, [emphasis added] though citations within any given article should follow a consistent style. 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." The widely used The Chicago Manual of Style, noted in that sentence, is among those that do not italicize websites. This RfC about a cite template appears to overreach to change MOS and mandate that Chicago Manual of Style not be used. Should not consensus for such a major site-wide MOS change, including throwing out Chicago Manual of Style, be reached here instead?

This RfC mandates a house style in contradiction to what WP:CITESTYLE states. Should it go to RfC Review or is this page the proper forum to discuss imposition of a house style? --Tenebrae (talk) 17:57, 9 September 2019 (UTC)

There is only a conflict if you use the template to formulate your citations... but if you use the old hand coded “<ref>citation text</ref>“ format, you are not limited. So... if you don’t like the style formatting used by the template, just don’t use the template. Blueboar (talk) 21:49, 9 September 2019 (UTC)

Thanks for responding, Blueboar! I did think of that. What I've found in practice is that editors invariably, eventually, take non-templated footnotes and put them into templates. I understand editors doing so, since it's for consistency. The result, though, is that in any practical sense, we're not allowed to use The Chicago Manual of Style or other styles that don't italicize websites.

So I'm thinking... What do editors here think of a new template called something like "cite organization" that would allow flexibility for either italicizing or non-italicizing websites? --Tenebrae (talk) 18:15, 10 September 2019 (UTC)

I'd say the last thing we need is more citation templates. Der Wohltemperierte Fuchs ^talk 18:37, 17 September 2019 (UTC)

So we should just give up and say, "OK, there's now one official house style that we all must use, no matter hwo eccentric and non-mainstream it is"? --Tenebrae (talk) 18:52, 17 September 2019 (UTC)

There is {{article style}} used in edit notices to make editors aware of a particular citation style. That template has support for CMOS. No doubt, editors will ignore the edit notice so {{article style}} isn't an absolute preventative, but it may be helpful. At en.wiki, cs1|2 is ubiquitous (at this writing 4,174,483 articles); editors at en.wiki and at many of the other-language wikis apparently like it.

For other editors: This topic has had some related discussion at my talk page.

—Trappist the monk (talk) 12:01, 18 September 2019 (UTC)

From the discussions over the period where cs1/2 had made the |website= parameters required, temporarily (a change since reverted) it seemed clear that there are certain classes of "websites" that should not be italicized, or more specifically, we use the name of the publisher as the key source parameter rather than the name of the website. This would include news stations (BBC, CNN, etc.) and entities like WHO and NASA. There are likely others but not limited to these classes. There, just using |publisher= should be sufficient since the website name is nearly always duplicative as there's no specialized site name, and that publisher parameter doesn't italicize the name. This appeared to be a common way to do it across ~100,000s of articles (based on error reports that got categorized when |website= was made mandatory), so that seems to be the solution most have settled on. --Masem (t) 13:58, 18 September 2019 (UTC)

Citations always mandatory?

The way I see it, citations should br for resolving disputes between editors. If you write something without providing a reference, it should be acceptable until some other editor takes issue. Requiring citations for all material in all circumstances discourages knowledgeable people from contributing. Like, say you're a surgeon of 20 years experience and you decide to add something on an article about surgery. You know the material like the back of your hand and it's common knowledge in the profession. But you don't add a citation, so somebody just deletes it and chastises you. You don't want to go through the hassle of digging out some books and articles. You're a busy man! You tell the other guy that you're a highly-qualified surgeon, but he says that's irrelevant, everyone on Wikipedia is equal regardless of their expertise. You throw up your hands in frustration and go back to lecturing college students -- you're actually paid to do that, and you get some respect.

This also ties in to the idea that Wikipedia is about the wisdom of the masses. Is it about the wisdom of the masses, or the books?

I would like the admins to clarify this a little for me, in case I get into an argument with an editor who deletes my work simply for lack of references, not because he knows better. Kurzon (talk) 16:29, 30 September 2019 (UTC)

This is something I face when editing Lexicanum, a wiki devoting to the game Warhammer 40,000. If I write something without a proper citation it will be automatically deleted by an admin and I will get reprimanded.

We cannot control what happens to you on a wiki which is not the English Wikipedia. Is this question about English Wikipedia or Lexicanum? --Izno (talk) 16:43, 30 September 2019 (UTC)

I sympathise with this perspective, but I do know that verification of data from secondary sources is one of the basic principles of Wikipedia. The 'inclusionist' part of me wants to say allow everything from someone's knowledge to be included, but that might open the floodgates to all kinds of inaccurate stuff which would mislead others. I am really torn by this issue, as I consider myself an inclusionist, but as a scholar I also know that we must maintain a basic level of verifiability. Ishel99 (talk) 03:19, 1 October 2019 (UTC)

Citations aren’t just about showing something is true; they also can show that a given subject matter is

WP:Biographies of living people where the default is to excise rather than tag unreferenced material; in general, I think, editors are given time to respond to requests for citations unless there are other issues with the text. Umimmak (talk

) 11:22, 1 October 2019 (UTC)

It could be argued that if you're too busy to provide a citation (even something as basic as a URL which can be cleaned up if necessary), then you're really too busy to be adding info to WP to begin with. Also, I think it should be common sense that editors can't typically or readily confirm other editors' credentials, so no, nobody should be allowed to add information to an article strictly on the grounds of their claiming that they're an expert; indeed, one would think that an expert in a field would understand the need to properly reference their work. DonIago (talk) 14:18, 1 October 2019 (UTC)

One of Wikipedia's rules is that the consensus of editors is ultimately what decides what does and doesn't get included. Not books. If two or more editors decide something is valid, then it is valid even if there is no reference. So why not a consensus of one? If only one editor is paying attention to an article and nobody else has an issue with what's written, then the material should be acceptable. Wikipedia also has a rule that all editors are equal regardless of their expertise, yet they are nothing if their contributions aren't backed up by a book written by an expert. This, to me, is a contradiction. Kurzon (talk) 15:35, 1 October 2019 (UTC)

Editors have the right to put unsourced information into an article. And other editors have the right to challenge that information if they believe it is

unverifiable

. As the easiest way to address even potential verifiability concerns is by providing a reference, is it not then better to be proactive and add a reference at the time you're adding information?

If editors are coming up with a consensus to the effect of "this information is unsourced and may not even be true, but we don't care" when anyone has expressed verifiability concerns, then they are acting counter to

WP:V

. Were I to see such a situation occur I would likely call for an RfC.

So yes, an editor acting on their own can add whatever they want to an article that nobody is taking issue with. But it would be naive to believe that there will never be some point in the future when another editor may take issue with it and ask for references, and at that point, it will ultimately be contingent on the editors who wish to retain the information to provide such references, as discussed at

WP:BURDEN

.

In my experience, the best option is to simply provide a reference rather than bicker over whether a reference is truly needed, which typically consumes everyone's time and energy just to typically conclude (predictably) that yes, a reference is needed. DonIago (talk) 16:00, 1 October 2019 (UTC)

That's fair. Kurzon (talk) 17:54, 1 October 2019 (UTC)

Citing other-language sources

I am interested in guidance for how to indicate that a source cited is in a language other than the language of the Wikipedia version being edited (I refrain from using the term 'foreign' languages since the term 'foreign' perhaps implies something which doesn't belong... I could be wrong though). I don't find anything about this so far despite searching (maybe just not using the best search terms?) so if anyone can please point me to anything that already exists that would be helpful. It would be good if guidance could even be provided on this main page about citing sources. Ishel99 (talk) 03:13, 1 October 2019 (UTC)

You may choose to use a

CS1/2 template in which case you should use |language=. You may choose otherwise to use manual citations, in which case you are free to do what you wish. --Izno (talk

) 04:40, 1 October 2019 (UTC)

Thank you for this! It works perfectly in the context I needed it for. Is there any reason I can't add that guidance to this page? Also, please excuse my ignorance, but what does the prefix 'para' achieve in the code? I just place |language= in my example and it worked. Ishel99 (talk) 07:07, 1 October 2019 (UTC)

{{para}} is used for discussing parameters used in templates; it just makes it the formatting look like “|language=” instead of just “|language=“. Umimmak (talk) 10:41, 1 October 2019 (UTC)

Thank you for the explanation about {{para}} . I would appreciate further advice here. I added the use of |language= to this page, and this was furher developed by another editor; but within a couple of hours the whole thing has been removed by yet another editor with the comment that 'this page is not a citation tutorial'. I fail to see the difference between what we added and many of the other guidelines provided here. Could someone please have a look and advise whether our contribution might be restored? Ishel99 (talk) 14:31, 1 October 2019 (UTC)

That was me. This page is not a cs1|2 citation template tutorial nor is this page specific to the particular citation style that is cs1|2. |language= is a templated citation style parameter use, most commonly in cs1|2 (also used in templates of other templated citation styles; the vcite templates come to mind). Other citation styles in use at en.wiki, templated or not, may or may not support the notion of annotating the source's language. |language= makes no sense in an untemplated citation style so should not be explicitly recommended here.

—Trappist the monk (talk) 14:55, 1 October 2019 (UTC)

(

Wikipedia:Manual of Style/China and Chinese-related articles#Citations). However, just saying something about a specific parameter in a specific set of templates is too specific. Some people don't even use citation templates; the information on this page should still be of use to them too. That was my understanding for why that was removed. Umimmak (talk

) 14:57, 1 October 2019 (UTC)

@Trappist the monk: @Umimmak: Thank you for the comprehensive explanations. It makes sense given the broader context to which you draw attention. However, I admit that so far I have been unable to find anything that does give me the advice I sought, in a more appropriate context. Does it exist anywhere? Ishel99 (talk) 15:39, 1 October 2019 (UTC)

|language= is documented on the cs1|2 template documentation pages (for example at Template:Cite book § Title) and there is also and example at Template:Cite book § Examples.

—Trappist the monk (talk) 15:51, 1 October 2019 (UTC)

@Trappist the monk: Thank you, those will be extremely useful. Just one more question: what would be the basic 'starting page' for citation in Wikipedia, which would clearly lay out all the options and then fork off to more detailed pages about each option? (I realize that, given the way Wikipedia works, there may actually not be any such page... but just wondered!) Ishel99 (talk) 00:29, 2 October 2019 (UTC)

You might try Help:Referencing for beginners and the pages linked from § See also, and Wikipedia:Citation templates.

—Trappist the monk (talk) 00:37, 2 October 2019 (UTC)

@Trappist the monk: Thank you so much, I appreciate the help... been in Wikipedia a long time, but still scratching the surface  :) Ishel99 (talk) 00:46, 2 October 2019 (UTC)

RFC regarding the scope of RfC regarding italicization of the names of websites in citations and references

Pursuant to a request by the closer:

There is a request for comment to definitively determine how widely the RFC Italics of websites in citations and references – request for comment should be applied. Please contribute.

—Trappist the monk (talk) 14:14, 6 October 2019 (UTC)

Internet Archive page links

Some news that may be of interest:

• Bailey, Lila (2019-10-30). "Fighting Misinformation Online". Internet Archive.
• Cory, Doctorow (30 October 2019). "The Internet Archive's massive repository of scanned books will help Wikipedia fight the disinformation wars". Boing Boing.
• Finley, Klint (2019-11-03). "The Internet Archive Is Making Wikipedia More Reliable". Wired.

I note that Katherine Maher has - not unreasonably - taken issue with the tone of Doctorow 's piece. [1]. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 12:02, 4 November 2019 (UTC)

See also the issue I have raised, at User talk:Cyberpower678#InternetArchiveBot page links about duplicate links being created. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 13:04, 4 November 2019 (UTC)

There's also this. Nortonius (talk) 13:45, 4 November 2019 (UTC)

Use of abbreviations and acronyms for citation sources

Quick question, is there any Wikipedia policy to using the proper name of a publication over an acronym or abbreviation. Is it preferable to use "

JJLA. I tend to lean to the proper name so as to avoid confusion as there may be many people unfamiliar with shortened form especially for those who are not from the country of the publication or who use English as a non-primary language.Patapsco913 (talk

) 11:28, 5 November 2019 (UTC)

worth mentioning: it's related to Special:LongPages, affecting dozens of refs each, resulting in unnecessary filesize blow-up. Tnnnbm (talk) 12:16, 5 November 2019 (UTC)

Rolling on-line news pages

If this has been raised before, I apologise, but couldn't see it in the archived discussions. It affects many current developing news articles.

References to rolling-news web pages soon become unusable as practical refs (like citing a book but no page number) - for example, in the current article

2019 London Bridge attack, this BBC page is already used nine times. For me at present that is 21 screen pages, but many of the uses of that ref will not be on that page because they have been shunted into subsequent pages like page 2 up to page 6 at present, each with about a dozen screenfuls each - scrolling through over eighty screen pages to find the actual source is just hopeless. And the total pages could easily be twice the size by this time tomorrow. The only improvements I have thought of so far are either deprecating the use of such pages in favour of more static news pages, or requiring the webpage to be archived at the time of use. Davidships (talk

) 00:47, 30 November 2019 (UTC)

See:

WP:NOTNEWS. Wikipedia is not really designed for reporting on current (breaking) news. It simply isn’t our purpose. The current state of the article should be thought of as a “first draft”... It will need some heavy editing and rewriting before it reaches its final state. Blueboar (talk

) 01:06, 30 November 2019 (UTC)

Yes, of course, that will happen in due course. But I'm asking about a particular kind of reference and whether there should be some guidance the use of online rolling-news pages because of their nature. I cannot see anything in

WP:NOTNEWS that helps with this. Davidships (talk

) 10:16, 30 November 2019 (UTC)

Rolling news pages? Well, first of all, please re-review

WP:NOTNEWS: "Wikipedia is not a newspaper". Particularly (bolding added): "Wikipedia should not offer first-hand news reports on breaking stories"; and: "Wikipedia considers the enduring notability of persons and event." "Breaking" news, by definition, is about events that have yet no "enduring". Such coverage is better done at Wikinews

.

I suspect the "nature" of your "online rolling-news pages" is transient, even ephemeral. (You didn't specify, but I would guess they are similar to the "crawler" often displayed at the bottom of the screen.) Don't forget that our core content policy of WP:Verifiability not only requires an attributable source, it is implied that such sources must be enduring, with some degree of permanance. If you describe what you saw on a screen (or anywhere), that is personal report (and what we call original research). If you want to cite it, you need to provide a reference where others can see it. If you're so close (in time) to an event that there is not (yet!) more permanent, usual forms of documentation then you are most likely following too closely. ♦ J. Johnson (JJ) (talk) 21:20, 30 November 2019 (UTC)

Thank you for the irrelevant lecture. They are not "my" pages and I wouldn't wish to cite them, precisely because they seem to be problematic for both readers and subsequent editors. They are essentially an online news media's chronological compendium of its reports on an event - but you would know that if you looked at the example I linked at the beginning. Davidships (talk) 02:04, 1 December 2019 (UTC)

Deprecating printers' names

Many books, especially author-published books, have the printer of the book on the title page. I haven't gone into the style manuals lately on this (did long ago), so call me on that if you feel it's appropriate, but I believe that including a printer in a citation should be deprecated as unhelpful at best (save for those few complex cases in which the printer of a particular copy or edition is important). It is most commonly misleading and often downright deceptive on the part of authors, making it seem falsely that the book was not self-published.

With really old books (the hand-typesetting age, when a printing press was a large investment, and publishers as distinct from bookstores or institutions did not exist) a printer on his own initiative would print his name in larger type than, or in place of, the name of the publisher, as a way to advertise his services. An example I am very familiar with is the 1605 first edition of Don Quixote, for which you may incorrectly find the printer Juan de la Cuesta given as the publisher, which he was not. The publisher — the one who hired and paid the printer, then sold the books himself and kept the profit if any — was the bookstore owner, in this case Francisco de Robles.

I feel strongly that this should be added to citation help files as a recommendation. Others?

P.S. I have used the masculine as I'm pretty sure there were no female book printers before 1950 except for widows, who would invariably describe themselves as "The widow of...". And very few after 1950.

P.P.S. If this is not the best place to post this, please tell me where it should go. deisenbe (talk) 12:57, 28 November 2019 (UTC)

Are you proposing a specific addition? If so, what? In very early books (before say 1550), the printer typically was the publisher. I can't say I'm aware of any common problem around this. Johnbod (talk) 13:50, 28 November 2019 (UTC)

I'm proposing that we recommend that book printers not be mentioned in citations, unless the printer was the publisher. I.e., publishers are worth citing. Printers are in most cases irrelevant. (Early printers like Manutius were indeed publishers, but many printers printed large quantities of government and religious items — much more lucrative. Literature or classics are viewed as important by us, but commercially they weren't very significant.) deisenbe (talk) 20:17, 1 December 2019 (UTC)

@

Help talk:CS1.  — SMcCandlish ☏ ¢

 😼  06:15, 3 January 2020 (UTC)

Numbered Periodicals and Nonstandard Volumes

Sometimes periodicals are numbered from first publication, and may or may not have volume and/or issue numbers. It gets confusing, and the "number" alias for "issue" in the citation template sometimes doesn't help. See this ref desk question. Here there is Bulletin #82, Volume 7, Issue 12; all on the title page.

1. "Documents pour l'histoire de Tahiti". Bulletin de la Société des Études Océaniennes. 7 (12). --just ignore the extra number, but #82 is probably the best way to find
2. "Documents pour l'histoire de Tahiti". Bulletin de la Société des Études Océaniennes No. 82. 7 (12). --best i think, maybe with parens (No. 82)
3. "Documents pour l'histoire de Tahiti". Bulletin de la Société des Études Océaniennes. 7 (82). --by following citation template
4. "Bulletin de la Société des Études Océaniennes numéro 82", Bibliothèque Numérique de l'Université de la Polynésie française --suggested on website, but omits lots

or - Num 27 - DU BON USAGE DE LA SOUFFRANCE - Automne 1994, numbers 1-51, 1982-2006, paginated by issue, not collected in volumes

1. Nobody, J. Q. (Autumn 1994). "An Article". Médiévales (27): 12. --page number confusing without volume
2. Nobody, J. Q. (1994). "An Article". Médiévales No. 27: 12. --even worse
3. Nobody, J. Q. (1994). "An Article". Médiévales. 13 (27) (published Autumn 1994): 12. {{cite journal}}: Check date values in: |publication-date= (help) --infer the volume
4. Nobody, J. Q. (1994). "An Article". Médiévales No. 27. 13 (2): 12. --infer the volume and issue

A somewhat related question, sometimes periodicals are collected, bound, and published in volumes that are just numbered (volume 1,2...), by a span of years (1948-51), or parts of years (Nos. 1-45, or October 1948 to July 1951). There are lots of combinations, how about [2], Newsletter #6, 1-64 bound, no pagination, confusing information about the volume. Using two citation templates is probably clearer for the reader, but might confuse automated tools looking at references? Can you get everything in a single template? use "volume" or "work" field? Can you get both periodical and volume publication dates in the template?

1. Missionary Sketches VI, London Missionary Society, July 1819 --ignore volume publication if any
2. Missionary Sketches VI, vol. 1, London Missionary Society (published 1840), July 1819 --add volume with "publication-year" for volume
3. Missionary Sketches VI, vol. 1, London Missionary Society (published July 1819), 1840 --add volume with "publication-date" for periodical, "year" for volume
4. Missionary Sketches VI, vol. Nos. 1-64, London Missionary Society, 1840 [July 1819] {{citation}}: |volume= has extra text (help) --more descriptive volume. "orig-year" for first pub., "year" for vol. pub.
5. "Missionary Sketches VI", Nos. 1-64, August 1820 to October, 1833, London Missionary Society, July 1819 --put volume info in "work" field
6. "Missionary Sketches", London Missionary Society, vol. 1, no. VI, July 1819 --go by template
7. Missionary Sketches VI, London Missionary Society, July 1819 collected in London Missionary Society (1840). Missionary Sketches Nos. 1-64, August 1820 to October, 1833. --use two templates

—

eric

16:32, 26 November 2019 (UTC)

For your first example, #2 (with the running number in the journal title) is certainly not best, because it makes it look like the number is part of the name of the journal and it makes other articles in other issues look like they are in other journals. If you're going to use parameters for the wrong information, just because of their formatting, why not avoid templates at all and just format it all manually? But I'd probably choose #1 (omit the running number), or maybe use the free-for-all parameter |at=No. 82:

"Documents pour l'histoire de Tahiti". Bulletin de la Société des Études Océaniennes. 7 (12). No. 82.

—David Eppstein (talk) 18:22, 26 November 2019 (UTC)

Thanks, "why not avoid templates"? I figure the ones who wrote the templates are better at this than me. I do want to get the right information in the right parameters, just don't know which is correct.—

eric

20:40, 28 November 2019 (UTC)

Another alternative:

{{cite periodical |title=Documents pour l'histoire de Tahiti |periodical=Bulletin de la Société des Études Océaniennes |number=82}}

"Documents pour l'histoire de Tahiti". Bulletin de la Société des Études Océaniennes. No. 82.

—Trappist the monk (talk) 21:46, 26 November 2019 (UTC)

I like that, works with volume, can't have an issue# at the same time, but #82 is probably the best way to find the source.—

eric

20:40, 28 November 2019 (UTC)

In "Bulletin #82, Volume 7, Issue 12", these are two conflicting numbering systems – "82" vs. "7(12)" – so just use one or the other. If you really want to include both, you could do something like |volume=7 |issue=12 [total issue 82]; or you could do |volume=7, |issue=12 |id=total issue 82.
PS: Peeps need to actually read the citation template documentation before making pronouncements about its parameters. The |at= parameter is not a "free-for-all parameter"; it is an alternative to |page=/|pages= for unnumbered source material. Ex.: |at=back cover, |at=tipped-in errata sheet; on an LP: |at=side B, track 3; in a documentary film: |at=01:23:14; in a database where you can't link directly to a specific record |at="Democratic Republic of the Congo" table; in a video game when citing dialogue: |at="Grin and Bear It" quest-initiating encounter with Temba Wide-Arm. Abusing |at= for things like issue numbers produces bogus citation metadata ("total issue no. 82" or whatever isn't something within the source, it's an alternative identifier of the source).
PPS: If you have a complex case, there is no reason at all you can't add a free-form note. e.g.: <ref>{{Cite book|[as many parameters as you can fill in accurately]}} Note: The material is found in a map inserted in but not physically attached to the book.<ref>

WP:Common sense is your friend. :-)
 — SMcCandlish ☏ ¢

 😼  06:46, 3 January 2020 (UTC)

Citing page numbers

WMDE has put their wishlist item for improving reference handling up on the beta wiki. You can see (and edit) a simple example of how it works at https://en.wikipedia.beta.wmflabs.org/wiki/Sandbox The main goal is to make it easier to cite different pages from a single book, without needing to repeat the entire citation. The new code for that is (currently):

<ref name="Example">Alice (2019) ''The Sun is Very Hot''.</ref>
<ref extends="Example">Page 1.<ref>

I like the way it groups all of the short citations with the full citation. I believe that feedback belongs at m:WMDE Technical Wishes/Book referencing. Whatamidoing (WMF) (talk) 22:00, 3 December 2019 (UTC)

It's a boondoggle, having a specious appearance of usefulness, but actually making "referencing" (citation) harder, and further ingraining some fundamental misapprehensions of citation practice. ♦ J. Johnson (JJ) (talk) 22:44, 3 December 2019 (UTC)

And meanwhile, very basic

WP:V problems like short references being used in articles without corresponding full references go undetected by the renderer, requiring user-written scripts to make them visible. How do regular editors help Wikimedia developers focus their limited resources on improvements that actually help maintain the integrity of the encylopedia? – Jonesey95 (talk

) 01:18, 12 December 2019 (UTC)

Add this to your common.js:importScript('User:Ucucha/HarvErrors.js');, errors are then easily picked up and corrected. Martin of Sheffield (talk) 11:48, 18 December 2019 (UTC)

Thanks for that.  — SMcCandlish ☏ ¢ 😼  06:48, 3 January 2020 (UTC)

Yes, I've been using that for a long time now and find it quite helpful. And maybe it's appropriate that readers don't get shown these errors quite as obtrusively as this script shows them. But it doesn't invalidate Jonesey's point, that the actual wiki developers don't seem to care about making this sort of thing more standard as part of what other editors see, rather than forcing the few of us who run the script to run around after everyone else who don't notice when they create these problems. —David Eppstein (talk) 07:33, 3 January 2020 (UTC)

And meanwhile, very basic

WP:V problems like short references being used in articles without corresponding full references go undetected by the renderer I'm really seriously confused here. This software development WMDE is undertaking will use a parameter extends which will require and target some other named reference. An error will be thrown by the software when it does not have such a target reference, i.e., you will get error checking in the software (and I have added such a case to the sandbox page that WAID linked). What are you saying the system should do instead? --Izno (talk

) 13:42, 3 January 2020 (UTC)

Website date and access date

We have users that regularly update facts from websites but dont update the dates and access dates, the link still works and the website has been changed but the dates that are used in the citation are clearly wrong. Do we just ignore this on regularly updated source pages or encourage users to actually update the dates? MilborneOne (talk) 16:32, 8 January 2020 (UTC)

The access date of a citation to a website is important - it is the equivalent of noting which “edition” of an old printed book is being cited. It tells the reader that “as of this point in time, the cited website supported what we say here” (and if it no longer does so, we need to look into why it no longer does so). Blueboar (talk) 19:18, 8 January 2020 (UTC)

It would be nice to apply the current date as a default, but there are problems with that. Templates are expanded upon access, not upon storage unless "subst"ed. Superficially we could require cite templates to be "subst"ed, but this then brings on other issues, including editability. I'd love to have an easy solution, but I can't see how the issue can be resolved technically. Martin of Sheffield (talk) 23:42, 8 January 2020 (UTC)

Templates are expanded upon access, not upon storage unless "subst"ed. I think that you are mistaken. When you click the Publish changes button, MediaWiki converts cs1|2 templates (and all of the other templates and text in the article) to html and stores the html – this is the cached copy of the article. When you 'access' an article as a logged-out reader, MediaWiki serves you the cached copy. When you access that same article as a logged-in editor, MediaWiki serves you the cached copy tailored to you (your skin and custom css, links to your talk page, your sandbox, your preferences). Prove this to yourself with your browser's view-page-source (usually a right click context-menu option in win xx operating systems) and ctrl-f search for your user name.

—Trappist the monk (talk) 00:12, 9 January 2020 (UTC)

I bow to your superior knowledge Trappist. The browser test you suggest though will not work, since the page has finished the server-side processing before it is delivered to your browser. If the pages are expanded upon storage, then is it possible to have a default date? As the acknowledged template expert can you comment please. Thanks, Martin of Sheffield (talk) 09:02, 9 January 2020 (UTC)

Of course the page has finished the server-side processing before it is delivered to your browser. The test was to show you the difference between what a logged-out reader sees and what a logged-in editor sees. Another thing you can do to prove to yourself that pages are cached is to 'View page source' and look at the NewPP limit report which lists Cached time: 20200107175101 (this from RS-485).

If I understand what it is that I think you want, templates can't do it. I think that I understand you to want the cs1|2 templates to automatically set |access-date= to the current date when a new {{cite web}} template is added to a page. Templates are processed anew each time MediaWiki processes the page. MediaWiki processes a page because an editor changed something; because something that is transcluded in the page has been changed; and perhaps for other reasons that I do not know about. Each time, the templates are reprocessed as if the page were a brand new page.

Your notion of substing is partially valid. You can write:

{{cite web |title=Title |website=Website |url=//example.com |access-date={{subst:date}}}}

But, caveat lector, substing does not work inside of <ref>...</ref> tags (a very old bug).

I think that leaves tools that create raw citation templates as the method for creating a default |access-date=; WP:RefToolbar will add |access-date=<current date> with the click of a button. I don't use ve, but I expect that tool will do something much the same.

—Trappist the monk (talk) 12:49, 9 January 2020 (UTC)

Hi Trappist, I think we are saying the same thing, but at a different level of detail. I initially said that it would be nice for {{cite web}} and friends to add a default, but then went on to explain that I thought it was technically impossible, just as you say above. The caching is a bit of a red herring, even if pages are cached for a short while (RS-485 had Cached time: 20200109151300 when I viewed it at 15:13) it is still done as part of Wiki's output mechanisms, not as part of the save/publish process. When I mentioned substing I had in mind {{susbt:cite web |title=Title |website=Website |url=//example.com}} but abusing templates in this way would, I'm sure, cause way more problems that it solves; hence my final comment to the OP of "I'd love to have an easy solution, but I can't see how the issue can be resolved technically". Please don't feel you need to reply to this (unless you want to), your time is far more valuable to WP. Regards, Martin of Sheffield (talk) 15:29, 9 January 2020 (UTC)

Time zone

When citing a source, what time zone should be used for the metadata, e.g. access-date? Is the editor's local time zone acceptable, or should UTC be used? - Indefensible (talk) 05:50, 4 January 2020 (UTC)

Since only the date is used, not the time, does it really matter? I'd always use the local date, but you could argue for Zulu = GMT = UTC if you thought that it was critical. Martin of Sheffield (talk) 10:57, 4 January 2020 (UTC)

For |access-date= (I initially missed that part), editor's local date is close enough for our purposes. Therefore there is no need to require conversion to UTC. If you want to go to that extra effort, there is no need to tell you that you can't. But the payback for your extra effort would be zero since it would be rarely done and there would be no way to indicate that you did it. ―Mandruss ☎ 12:11, 4 January 2020 (UTC)

Thanks for the feedback. Since web sources and the Wayback Machine may be timestamped to the next day in advance of the editor's local time, it can look odd that the access-date and date and/or archive-date do not match for a ref, as if the editor is accessing information from the future. - Indefensible (talk) 07:27, 11 January 2020 (UTC)

Citing sermons

Is it possible to cite sermons for a biography of a religious person and preacher to verify the beliefs he espoused? Those sermons are audio files which have been recorded and can be accessed only with permission of the religious group in question to which he belongs. Thanks in advance. In Citer (talk) 22:26, 12 December 2019 (UTC)

These are self-evidently

talk

) 23:10, 12 December 2019 (UTC)

Depends on what they're cited for; see

WP:PRIMARY. JMF is correct that they're obviously primary sources, like their blog or Twitter posts, or an interview of them on The Tonight Show, or an op-ed on their politics that they got published in The Times.  — SMcCandlish ☏ ¢

 😼  06:08, 3 January 2020 (UTC)

In Citer, recordings that require special permission to access are not usable on Wikipedia at all. They are not considered Wikipedia:Published#Accessible, and therefore citing them will always violate WP:Verifiability rules.

There is nothing unique about sermons in this situation. You can use {{Cite speech}} for any type of speech, whether it's religious, political, ceremonial, or anything else. Sermons, in particular, can get cited as speeches (e.g., if there's a video available somewhere) or as print material (if it was printed on paper, which used to be a popular thing). But nothing can be cited if the general public has no realistic chance of getting its hands on it. We are quite liberal in what that means; we don't mind if you have to fly halfway around the world or pay thousands of dollars for a source, but access has to be offered to the anyone and everyone, not just to selected scholars or members of a religion. WhatamIdoing (talk) 03:05, 20 January 2020 (UTC)

Linking to Google Books

In I Wonder as I Wander someone has linked to Google Books using page numbers, but the pages of the book are not there. I am wondering what the best way is to shows there is a link to Google Books, even if it just has information about the book and not actual pages.— Vchimpanzee • talk • contributions • 17:42, 13 January 2020 (UTC)

If memory serves, SarahSV and User:Calliopejen1 have used this style.

Whether you can see the linked pages at Google Books depends upon (at least) your location, and access may change over time. In that article, most of the linked pages are visible to me. WhatamIdoing (talk) 03:14, 20 January 2020 (UTC)

WhatamIdoing I am referring to reference no. 2. This is the link that doesn't show the page numbers when you go to Google Books, but rather general information about the book. How, then, should Google Books be used?— Vchimpanzee • talk • contributions • 20:20, 22 January 2020 (UTC)

I've mostly stopped doing it because the links disappear so often. Vchampanzee, you can link to Google Books in several ways. For example with a page link (this is if you're writing refs manually):

Bryant, Chad (2007). ''Prague in Black: Nazi Rule and Czech Nationalism''. Cambridge, MA: Harvard University Press, p. [https://books.google.com/books?id=NOCOVmwJFMMC&pg=PA167 167].

or with a more stable generic link :

Bryant, Chad (2007). ''[https://books.google.com/books?id=NOCOVmwJFMMC Prague in Black: Nazi Rule and Czech Nationalism]''. Cambridge, MA: Harvard University Press, p. 167.

If you're using a template, you can add the generic link to url= or the page link to p=. SarahSV ^(talk) 05:03, 20 January 2020 (UTC)

Curiously in this example, for me, the linked page on Google books doesn't allow preview, but [3] (to a slightly different edition of the same book) works. As WhatamIdoing says, whether a Google Books preview link works is not something you can say "yes" or "no" to; in general it will depend on who is doing the preview, where they're coming from, and when they try it. For this reason, among others, I prefer Internet Archive book links when possible. —David Eppstein (talk) 20:23, 22 January 2020 (UTC)

I have changed the link in the article to the one you posted, but it doesn't give me page 201 for some reason.— Vchimpanzee • talk • contributions • 20:45, 22 January 2020 (UTC)

Reprints

If a book is shown as "First published 1988, reprinted in paperback 1997", with no indication of any revision, how should the date be cited? The guidelines do not appear to cover this situation, as orig-year= seems to be intended for older publications. Dudley Miles (talk) 10:24, 4 February 2020 (UTC)

Per

WP:SAYWHEREYOUREADIT

, if the book that you consulted is the paperback reprint, use the reprint date. You might also set |edition=reprint.

—Trappist the monk (talk) 12:30, 4 February 2020 (UTC)

Location

I would like to add something about how to write locations in citations. There was a disagreement about this recently, and the guideline says nothing about it, except "Citations for books typically include ... place of publication". Options:

• Kürbis, Nils (2019). Proof and Falsity: A Logical Investigation. Cambridge: Cambridge University Press.
• Kürbis, Nils (2019). Proof and Falsity: A Logical Investigation. Cambridge, England: Cambridge University Press.
• Kürbis, Nils (2019). Proof and Falsity: A Logical Investigation. Cambridge, UK: Cambridge University Press.
• Kürbis, Nils (2019). Proof and Falsity: A Logical Investigation. Cambridge University Press.

---

• Rawls, John (1972). A Theory of Justice. Cambridge: Harvard University Press.
• Rawls, John (1972). A Theory of Justice. Cambridge, MA: Harvard University Press.
• Rawls, John (1972). A Theory of Justice. Cambridge, Mass: Harvard University Press.
• Rawls, John (1972). A Theory of Justice. Cambridge, Massachusetts: Harvard University Press.
• Rawls, John (1972). A Theory of Justice. Harvard University Press.

The latest

Chicago Manual of Style

(17th edition) prefers the two-letter postal codes for US and Canadian publications (p. 583), e.g. Cambridge, MA. It omits postal codes for major cities, adding them if the location is not well-known or might cause confusion (pp. 813–814). That means (examples from pp. 799–814):

• For the US: Baltimore, Berkeley, Boston, Bloomington, Chicago, New Orleans, New York, but Princeton, NJ; New Haven, CT; and always Washington, DC.
• For the UK: Harmondsworth, UK, but London, Cambridge, Oxford.

When the state is part of the publisher's name, it needn't be added again, so Chapel Hill: University of North Carolina Press. And use English names: Munich, not München.

I thought I would add a brief summary of the above to the section "What information to include" for books, making clear that editors can choose whichever style they prefer, with the usual advice about consistency and, in cases of dispute, deferring to the style used by the first major contributor. We should probably also say something about how to find the location; I usually look on WorldCat. Any thoughts would be appreciated. Pinging Ealdgyth, who has handled a lot of source reviews. Also pinging the editors who were involved (directly and otherwise) in the recent disagreement about this: Victoriaearle, Ceoil, Gog the Mild, Jonesey95, Finnusertop, Serial Number 54129, Izno. SarahSV ^(talk) 01:02, 12 December 2019 (UTC)

Another ping: SMcCandlish. SarahSV ^(talk) 01:08, 12 December 2019 (UTC)

Thanks for the ping. A friendly suggestion: If you expect editors to follow a citation style that conflicts with the normal guidance provided in the Manual of Style, in this case

MOS:POSTABBR (which until yesterday's edit by this thread's original poster said not to use postal abbreviations), please post a comment somewhere in the article explaining the non-MOS-compliant citation style that is in use. You can't expect editors to be mind-readers, or to understand arcane abbreviations that are not wikilinked. As it says in the lead of that MOS page, Always consider whether it is better to simply write a word or phrase out in full, thus avoiding potential confusion for those not familiar with its abbreviation. – Jonesey95 (talk

) 01:13, 12 December 2019 (UTC)

And at bare minimum use {{

MOS:ABBR as a general abbreviation-handling matter. (In other words, a "screw the readers, make a few specialist editors happy instead" questionable IAR argument to defy MOS:POSTABBR in order to comply with an external citation style, doesn't magically transmogrify into immunity from everything in MOS:ABBR.) All that said, it would really make more sense to open a discussion on the article's talk page about changing to a citation style that's less "reader-hateful".  — SMcCandlish ☏ ¢

 😼  16:27, 16 January 2020 (UTC)

(edit conflict) I believe those are called "postal abbreviations", rather than postal codes (to differentiate it from the numeric postal code).

It might also be useful to ping User:Trappist the monk, who coordinates {{cite book}}. He's likely to know if anyone runs AWB to strip out "unnecessary" locations. WhatamIdoing (talk) 01:15, 12 December 2019 (UTC)

Hi WhatamIdoing, the CMOS calls the two options "two-letter postal codes" (IL, MA) and "the conventional state abbreviations" (Ill, Mass). Jonesey95, re: the MoS, the advice to avoid abbreviations was added recently without discussion, and it conflicts (I believe) with what most editors do. This guideline is the place to add advice about how to write citations. SarahSV ^(talk) 01:28, 12 December 2019 (UTC)

Postal abbreviation (disambiguation) is where you'll find links to the lists of those abbreviations.

The style guides you've cited are written for a primarily American audience, which is likely to know what most of those abbreviations mean. I wonder whether people from other countries will see "DE" and think

Deutschland instead of Delaware, or see "IL" and think Israel instead of Illinois. I think that about 40% of the US two-letter state abbreviations are also the two-letter abbreviation for another country. WhatamIdoing (talk

) 06:41, 12 December 2019 (UTC)

New Hart's Rules too (Oxford University Press, 2014, p. 358): "It is not necessary to include the name of the country, but US state abbreviations can be included, particularly if the locations are globally less well known (Boca Raton, FL). Be consistent about the abbreviation system used (FL or Fla, etc)." SarahSV ^(talk)

06:55, 12 December 2019 (UTC)

I think it's best to spell out the names of US states in full, per

WP:NOTPAPER and to avoid e.g. the question of whether CA means California or Canada. I don't think it's reasonable to expect non-Americans to recognize the abbreviations of the state names. Or for that matter even for Americans to remember whether IA is Iowa or Indiana and whether AK is Alaska or Arkansas. —David Eppstein (talk

) 07:24, 12 December 2019 (UTC)

Yeah, the idea that Americans all have this memorized is false. Lots and lots of Americans can't reliably remember the postal abbreviations of the M states (MA, MD, ME, MI, MN, MO, MS, MT). The US system is pretty much just random rather than following a clear pattern (technically, it's a commingling of several conflicting patterns, plus some completely arbitrary choices), and with many of them being innately ambiguous. It would have been much more sensible to pick unique letters where possible, e.g. MY for Maryland, MG for Michigan, MP for Mississippi, and so on, obviously use the first letter of each part for two-part names (SC, NM), and otherwise stick to a single regular pattern (e.g., first two letters, or if still ambiguous, first and third, and so on). Never mind that there are also already the late-added codes MH for Marshall Islands and MP (huh?) for Northern Marianas, two US territories; nearly no one knows these, but they're just as "official" as CA for California and NY for New York. Next, some other systems aren't much better; e.g. traditional British abbreviations sometimes don't make any sense in relation to the full place names in Modern English (viz. Oxon and Hants for Oxfordshire and Hampshire). Worse, many traditional and even some official abbreviation/code system are not consistent even within the same country (sometimes even for quite recent stuff, like the two conflicting codes for the Nunavut territory of Canada; NU is the postal one, while NV has been used for some other official contexts despite the trans-national North Am. postal conflict with Nevada; then a Canadian style guide says to eschew the official two-letter codes, use traditional ones like Que., and in that system do not abbreviate Nunavut at all [4]). There are innumerable conflicts between traditional but not standardized abbreviations and official codes (Calif. and CA, PEI and PE, etc.), with the former remaining more common in some countries, and also many conflicts between different traditional abbreviations (Ca., Cal., Cali., Calif.; Mx, Middx, Midx, Mddx, Msx, M'sex, etc.). Meanwhile, most jurisdictions around the world just use numeric postal codes. The idea that any of these these things are generally intelligible across our readership just isn't tenable. Trying to force them to be permissible would just lead to more

MOS:POSTABBR already says. The only reason this current debate is open is because some editors don't want to follow POSTABBR (just to defy MoS for the hell of it, or to save a few keystrokes, or to obsessively follow some off-site and basically Wikipedia-irrelevant citation stylesheet), and are thus artificially re-manufacturing the conflict that guideline already resolves for us.  — SMcCandlish ☏ ¢

 😼  17:18, 16 January 2020 (UTC)

I think that you give me credit for knowing more about the wide world of citation use at en.wiki than I actually do. Unless someone is doing something that breaks cs1|2 templates in ways that put articles into categories that I'm currently monitoring, I will likely never notice.

At the moment, I don't think that I have an opinion about this topic. When / if this discussion impacts how cs1|2 operates then I will likely have an opinion.

—Trappist the monk (talk) 01:43, 12 December 2019 (UTC)

In my experience the current usage here is not to include the location at all. The second line of this very content guideline reads:

Ritter, R. M. (2003). The Oxford Style Manual. Oxford University Press. p. 1.

ISBN 978-0-19-860564-5

.

I recommend not trying to set a standard for this and allowing all forms. We are comfortable with a variety of reference styles here. And don't try to enforce consistency within an article. For older books it can be necessary to specify a location if several publishers have the same name, but that shouldn't enforce a style on the whole article. For newer books the ISBN itself will link to enough information to tell which publisher is which. StarryGrandma (talk) 01:32, 12 December 2019 (UTC)

StarryGrandma, I agree about not requiring it, but most publishers do add location, and editors often do, so it makes sense to offer advice about it in this guideline. Also, I would say internal consistency is always expected. But the proposed edits would make clear that adding location, or adding it in any particular form, is not required. SarahSV ^(talk) 01:42, 12 December 2019 (UTC)

StarryGrandma, APA style 6th ed. does call for including the location, but only for books, not journals. Since it tends to emphasize journals as sources, it's easy to miss. As for Oxford, I recall seeing the guidance in some style manual that if the city was included in the name of the publisher, it wasn't necessary to give the location. But checking the APA and Chicago, I don't see that guidance. Maybe I'll eventually remember where I saw it. Jc3s5h (talk) 03:14, 12 December 2019 (UTC)

Jc3s5h, it could be "New York: Cambridge University Press", so the university name alone won't necessarily give you location. The best thing is to look at the copyright page of the book itself, including on Google Preview if you don't have a copy. If that isn't clear (and it sometimes isn't), then look it up on WorldCat. SarahSV ^(talk) 03:26, 12 December 2019 (UTC)

We do need a summary of the basic principles, but

WP:TALKFORK problem. In this case, I'm trying to prevent CREEP/POLICYFORK between MOS:INFOBOX and WP:CITE, at the cost of centralizing discussion at another page, at least for one round.
 — SMcCandlish ☏ ¢

 😼  04:17, 13 December 2019 (UTC)

It should be here because this is the guideline about how to write citations. The point is simply to describe the options. Locations aren't required, and if people want to add them, they can write them in the way they prefer. SarahSV ^(talk) 04:21, 13 December 2019 (UTC)

The approach this guideline has always taken is to allow an article to follow an outside style guide like Chicago or APA, or an idiosyncratic style for a particular article. If guidelines on locations were created that disagree with outside style guide, it would be a change in direction for this guideline.

I personally don't have much sympathy for idiosyncratic styles for each article, but it is convenient to be able to follow outside style guides, since various software and websites can help in creating such citations (not to mention editors who are already familiar with them). Also, some publications provide suggested citations for their content in several outside citation styles. Jc3s5h (talk) 04:57, 13 December 2019 (UTC)

• Why must we attempt to micromanage everything. There is nothing wrong in having a citation that has "Los Angeles, CA", rather than the longer "Los Angeles, California", as long as the citations in an article are all consistent. Can anyone give any actual benefits to the use of the additional eight letters? I mean real benefits, not nonsense like 'DE might get confused for Germany'? - SchroCat (talk) 22:50, 16 December 2019 (UTC)

• Personally I think locations in cites are generally useless, but if they are to be included, I can see the argument (for places less well known than Los Angeles) that two- or three-letter state/province abbreviations are useless to people not from that country. Nikkimaria (talk) 01:59, 17 December 2019 (UTC)

• I'm not so sure on that: I'm not American and either know or can accurately guess most of them. Many people won't know or care, but it's not there for instant knowledge, it's there to help find the edition of the book used, so if it's not 100% clear in the first place, it will be when the edition of book is searched. I'll keep using locations because there are still some books published that differ between publication location, although the ISBN/OCLC does make it a bit moot. - SchroCat (talk) 06:59, 17 December 2019 (UTC)

• Yeah, and it's also simply a part of the standard citation format in most citation formats. That is, WP:CITE can't have it both ways: Either every known citation style is permissible, and editors can insist on maintaining a particular one (with various features of it being mandatory in that particular system) at a particular article unless trumped by an RfC or other clear discussion on the talk page; or WP shouldn't be using external citation styles at all, and should instead normalize its own (probably CS1) across all articles, with internal control over whether it includes things like |location=, in what form, and under what conditions. The latter is obviously far more sensible and is what the majority of editors do anyway, but the "give me [insert academic organization here] citations or give me death" people have a chokehold on WP:CITE, so we'll probably be stuck with the current mess indefinitely. (I'm not sure how to generate stats on this, but I would wager that over 90% of our completed citations – i.e. not including drive-by pastings of URLs and other broken stuff – are in CS1 format, and it clearly dominates not just in raw numbers but across categories, with only a few narrow topical exceptions.) As for the "why 'Mississippi' instead of 'MS'?" question, it's because US postal state abbreviations are often meaningless to non-US readers (same goes for non-US codes and abbreviations – do you know what NT, NU, Hants, Rads, SXM, GGY, and FR-MAY all mean?). Postal and ISO 3166 codes for subnational jurisdictions exist for particular purposes and are not common global knowledge, not really intended for bibliographic purposes, can change over time, and can conflict between different standards. Places' full names change much less often and don't require any specialized knowledge to interpret. Because
  WP:NOT#PAPER, we really don't care that it requires a few extra characters to provide clearer information; column space is not at a premium here. On "for places less well known than Los Angeles": We've already got a standard that we do not need to provide subnational jurisdiction names or even the country for world-famous places like L.A., Paris, Tokyo, etc. Just |location=Munich will suffice. PS: In an encyclopedic context, location information can actually serve additional purposes beyond locating a particular edition; one often sees this come up in articles on ethnic strife and the like. E.g., in an article on countries formerly part of Yugoslavia, an overabundance of, say, Serbia-published sources in an article on a Croatian subject can reveal an untoward bias that might otherwise be invisible to a non-expert reader or editor.  — SMcCandlish ☏ ¢
   😼  06:04, 3 January 2020 (UTC)

Locations may matter because the edition may be counted from 1st per country, the year may be the same, and the pulisher may be part of the same international publishing house, but the printer may differ and hence the pagination may differ. With modern publications the ISBN can be used to identify the book, but for older books and as a cross check that the isbn is correct, location (eg London or New York) is useful. However I agree with SchroCat's post (22:50, 16 December 2019), as I see no need for consistecy and MOS guidence on this issue. "Americans to remember whether IA is Iowa or Indiana and whether AK is Alaska or Arkansas" I only need the work it out if I need it to access the source. However if "IA" works in a search to find where I can access the source, I don't care if it is short for Iowa or Indiana. I would only care if it made searches easier on the state name rather than the abreviation. What I find most amusing is the proposal to use Salem, {{abbr|MA|Massachusetts}} as the same reference may also include {{

issn}} and {{Webarchive}}, yet one can not wrap the information up into {{citation}} because of a 15+ year old argument about "citation templates". -- PBS (talk

) 14:38, 15 February 2020 (UTC)

Not A Number?

At https://www.nytimes.com/2011/02/06/us/06bclocalintel.html?searchResultPosition=10, there's a great footnote: A version of this article appears in print on Feb. 6, 2011, Section A, Page NaN of the National edition with the headline: Children’s Fairyland. I so want to cite this with

 {{cite newspaper
 ...
 | page=NaN
 }}

-- RoySmith (talk) 17:17, 8 February 2020 (UTC)

Are you wanting to cite something you actually read on a page that carried a "number" of "NaN"? Or are you trying to cite something that someone else cited? Or perhaps you want to cite an instance of "NaN" appearing as a page number?

Presumably you know that "NaN" is, indeed, "Not a Number" itself, but used where a data value is undefined or unrepresentable as a number. Its appearance usually results from corruption in someone's data; which is to say, it's an error of data, not of pagination. So was a page number on an issue of the NYTimes garbled? Or was it just in the citation in the footnote? In the former case determine the actual page number and use that. (Possibly in square brackets to show that is an emended value, but cs1 might choke on that.) Use of |page=NaN says that the value is "undefined or unrepresentable", not that there is a page whose number is unrepresentable. ♦ J. Johnson (JJ) (talk) 19:51, 8 February 2020 (UTC)

I think RoySmith knows all that. It's just that it's so tempting to cite the page number as NaN anyway, for its humor value. —David Eppstein (talk) 20:24, 8 February 2020 (UTC)

Ah! And perhaps my sense of humor has been so thin of late? :-) ♦ J. Johnson (JJ) (talk) 22:04, 8 February 2020 (UTC)

Indeed. -- RoySmith (talk) 22:50, 8 February 2020 (UTC)

NaN ougt to be a no no. -- PBS (talk) 14:40, 15 February 2020 (UTC)

Proposed shortcuts for "Additional annotation" section

It took me a while to find the actual section "Additional annotations" in an article dispute about footnote quotations. It seems to reasonable to add the shortcuts

WP:QUOTE, which currently points to an essay, since this is a WP content guideline (but I guess that might require renaming the essay). — MarkH₂₁^talk

20:20, 3 February 2020 (UTC)

Been about three weeks, so going ahead and adding the shortcuts. — MarkH₂₁^talk 11:30, 24 February 2020 (UTC)

What format for citing an album's liner notes

Is there a recommended format to cite an album's liner notes? I see Wikipedia talk:Citing sources/Archive 9# Citing liner notes, but the link there given is a red one, and I can't find the referred-to text, even accounting for page moves. I'm specifically looking at fixing up Nikita Magaloff (you can see my take so far at [5]). TJRC (talk) 22:55, 1 March 2020 (UTC)

Yes, see Template:Cite AV media notes. StarryGrandma (talk) 23:04, 1 March 2020 (UTC)

Thanks! TJRC (talk) 23:36, 1 March 2020 (UTC)

If you are using {{citation}} rather than {{cite XYZ}} see Carraig Aonair (The Lone Rock) for an example that seems to work satisfactorily. Martin of Sheffield (talk) 08:09, 2 March 2020 (UTC)

Citing a photograph and photographic editorial database?

I don't think this has come up before, and this case certainly isn't addressed in the article or with a template (exact perhaps with the AV template), but I see no reason why it shouldn't actually be permissable.

When information is only available through a photograph and the caption provided by a photography agency or other editorial company dealing in the supply of photographs to the media, could this be considered a citable source? To give you an example, here is an image that has a caption with information, from a reliable and respected photo agency, that may not be possible to find elsewhere through a similarly reliable source. Is there any reason this couldn't be used as a source? I'd consider a caption from a photojournalist to be as worthy as an article from a journalist, but perhaps others disagree? KaisaL (talk) 23:26, 3 March 2020 (UTC)

I think it's reasonable to cite a photo caption from a reliable agency, but of course a more conventional text source is preferred. Captions are often short, may lack context, and may lack details on where the information came from. So perhaps not as worthy as an article from a journalist, but a usable substitute if nothing else can be found. — MarkH₂₁^talk 23:29, 3 March 2020 (UTC)

I can understand that. I feel it's useful for basic facts that might be unavailable elsewhere, for example, as photo libraries often have quite good retention compared to news articles. The example I gave, for instance, confirms the dates of a test and that Pedro de la Rosa tested the specific car, which is minor information that's only available from forums, tweets, and so on otherwise. I don't know if this sort of citation needs its own template (probably not), but I do feel it opens up a new avenue for sourcing snippets of information. KaisaL (talk) 23:50, 3 March 2020 (UTC)

Amama Ede oballa Nsukka

Amama Ede oballa Nsukka, is one of the largest village in Ede oballa also highly populated village, it have up-to 7 different Otobo's with one 'onyishi' as the head of the village. Tochi Samson (talk) 10:06, 13 March 2020 (UTC)

@Tochi Samson: – I think you may have posted this in the wrong place, it has nothing to do with citing sources. Martin of Sheffield (talk) 10:41, 13 March 2020 (UTC)

broken harv link reporting

Please see the discussion at Module talk:Footnotes § broken harv link reporting where a broken harv-link reporting scheme is proposed.

—Trappist the monk (talk) 17:46, 16 March 2020 (UTC)

Clarification

Is

WP:CITESTYLE in this guideline intended to apply to the style of inline referencing used within an article, for example, if an article mixes both <ref>...</ref> and {sfn}? Factotem (talk

) 16:20, 19 April 2020 (UTC)

No, displayed style in the general case (see also). You might prefer to use {{sfn}} only for citations requiring book references and use {{cite web}} directly in the article and that would be okay. --Izno (talk) 19:18, 19 April 2020 (UTC)

String of tweets

There is an erudite debate at

talk

) 10:54, 21 April 2020 (UTC)

Listing newspapers, not issues, but just the names of the papers

Please see Talk:Isaac C. Smith#Listing newspapers and journals (entire runs, not individual articles or issues). -- JHunterJ (talk) 11:57, 13 May 2020 (UTC)

Clarification on page ranges might be useful

I think we could use a clarification on page numbers to cite for books, book chapters, journals etc. Recently I had discussion with an editor who thinks that this guideline requires, for example, that when citing an academic article it is required to add page numbers for a specific page (not just the page range of the article in a journal), and if it is not done, the citation should be marked with {{page needed}}. Are they going too far or is it indeed what this guideline recommends? (For the record, I see nothing wrong with being specific when it comes to page numbers, but I think it is too far to tag page needed for journal citations as it goes above and beyond our standards, including my reading of this guideline). Thoughts? --_{Piotr Konieczny aka Prokonsul Piotrus| reply here} 03:42, 30 April 2020 (UTC)

WP:CHALLENGE. More generally, what you cite should correspond to what you say. If you're trying to summarize Alice's latest book in a single sentence, then you should cite the whole book (without any page numbers). If you're pulling a single, isolated factoid out of a journal article, then it might be better (though by no means common) to cite the single page that it appears on. WhatamIdoing (talk

) 19:13, 14 May 2020 (UTC)

Archiving or future-proofing subscriber-only articles

Please see Help talk:Using the Wayback Machine#Archiving or future-proofing subscriber-only articles. — CR4ZE (T • C) 03:38, 17 May 2020 (UTC)

Is there a list of {{

cite

}} types?

I know of several {{

cite}}. Is there such an article? Shmuel (Seymour J.) Metz Username:Chatul (talk

) 02:59, 24 May 2020 (UTC)

@Chatul:..Wikipedia:Citation templates.--Moxy 🍁 03:02, 24 May 2020 (UTC)

The main ones are in Help:Citation Style 1. There are a lot of special-purpose ones not listed there, though. —David Eppstein (talk) 04:26, 24 May 2020 (UTC)

There is a link to Category:Citation Style 1 specific-source templates. Thanks. Shmuel (Seymour J.) Metz Username:Chatul (talk) 22:53, 24 May 2020 (UTC)

Are you referring to Wikipedia:Citation templates#Examples? That's not a complete list. Shmuel (Seymour J.) Metz Username:Chatul (talk) 22:53, 24 May 2020 (UTC)

Initials with/without full-stop (period)

Is there any right way after putting initials? I always used initials with full-stops. In fact I thought that was what one of the citation styles said, but I can't remember which – Vancouver? Eg. Barbara K Lewalski or Barbara K. Lewalski
Auntie Kathleen (talk) 13:29, 27 May 2020 (UTC)

@

talk

) 13:34, 27 May 2020 (UTC)

Long citations

I myself like using #{{cite... rather than *{{cite... which means the list of long citation as a numbered list, not just an unordered list
E.g.

1. Bennet, John (Jan 1792). Letters to a young Lady. Vol. 11. Philadelphia: American Museum. {{cite book}}: Invalid |ref=harv (help)
2. Blank, Antje (2003). "Charlotte Smith". In Todd, Janet; Cook, Daniel; Robinson, Daniel (eds.). The Literary Encyclopedia. Vol. Volume 1.2.1.06: English Writing and Culture of the Romantic Period, 1789–1837. Archived from the original on 13 October 2003. {{cite encyclopedia}}: |volume= has extra text (help); Invalid |ref=harv (help); Unknown parameter |editor1link= ignored (|editor-link1= suggested) (help)
3. Craciun, Adriana (2005). British Women Writers and the French Revolution – Citizens of the World. Palgrave Studies in the Enlightenment, Romanticism and Cultures of Print. Palgrave Macmillan.
   ISBN 978-1403902351. {{cite book}}: Invalid |ref=harv (help
   )
4. Curran, Stuart (1993). The Poems of Charlotte Smith. Women Writers in English 1350–1850.
   ISBN 978-0195083583. {{cite book}}: Invalid |ref=harv (help
   )
5. Curran, Stuart, ed. (2005). The Works of Charlotte Smith (14 vols). Vol. Part I: vols 1–5. Volume Editors: Adriana Craciun, Stuart Curran, Kate Davies, Elizabeth Dolan, Ina Ferris, Michael Gamer, M O Grenby, Harriet Guest, Jacqueline Labbe, D L Macdonald, A A Markley, Judith Pascoe, Judith Stanton and Kristina Straub. Pickering & Chatto / Routledge.
   ISBN 978-1851967896. {{cite book}}: Invalid |ref=harv (help
   )

versus

• Bennet, John (Jan 1792). Letters to a young Lady. Vol. 11. Philadelphia: American Museum. {{cite book}}: Invalid |ref=harv (help)
• Blank, Antje (2003). "Charlotte Smith". In Todd, Janet; Cook, Daniel; Robinson, Daniel (eds.). The Literary Encyclopedia. Vol. Volume 1.2.1.06: English Writing and Culture of the Romantic Period, 1789–1837. Archived from the original on 13 October 2003. {{cite encyclopedia}}: |volume= has extra text (help); Invalid |ref=harv (help); Unknown parameter |editor1link= ignored (|editor-link1= suggested) (help)
• Craciun, Adriana (2005). British Women Writers and the French Revolution – Citizens of the World. Palgrave Studies in the Enlightenment, Romanticism and Cultures of Print. Palgrave Macmillan.
  ISBN 978-1403902351. {{cite book}}: Invalid |ref=harv (help
  )
• Curran, Stuart (1993). The Poems of Charlotte Smith. Women Writers in English 1350–1850.
  ISBN 978-0195083583. {{cite book}}: Invalid |ref=harv (help
  )
• Curran, Stuart, ed. (2005). The Works of Charlotte Smith (14 vols). Vol. Part I: vols 1–5. Volume Editors: Adriana Craciun, Stuart Curran, Kate Davies, Elizabeth Dolan, Ina Ferris, Michael Gamer, M O Grenby, Harriet Guest, Jacqueline Labbe, D L Macdonald, A A Markley, Judith Pascoe, Judith Stanton and Kristina Straub. Pickering & Chatto / Routledge.
  ISBN 978-1851967896. {{cite book}}: Invalid |ref=harv (help
  )

Is there any actual reason for one or the other? Auntie Kathleen (talk) 13:23, 28 May 2020 (UTC)

The way citation templates are usually used, a passage in the text of a Wikipedia article (Coordinated Universal Time, for example) that is supported by an inline citation has a number in brackets next to it, like this: ^[43] (but it's blue). Clicking on the number brings you to the list of citation, numbered in the order in which they appear in the article (more or less).

In articles with many citations, the citation associated with an endnote number may be brief, such as "43. McCarthy & Seidelmann 2009, p. 54." Clicking on this brings you to the full bibliography, with an entry like this:

• ISBN 978-3-527-40780-4. {{cite book}}: Invalid |ref=harv (help
  )

It would be unconventional and perhaps confusing to have two numbered lists at the end of the article, the end notes in order of appearance in the article, and the bibliography in alphabetical order. Jc3s5h (talk) 13:39, 28 May 2020 (UTC)

As well as unconventional it would be somewhat pointless. A numbered list sort of implies one thing after another:

1. Boot computer
2. Log on
3. Go to Wikipedia

Whereas an unnumbered list adds no such constraint. In your example you don't have to read Bennet (1792) before going to Blank (2003) if the information is in Blank. Martin of Sheffield (talk) 16:37, 28 May 2020 (UTC)

Discussion about harmonizing Citation Styles 1 and 2

 You are invited to join the discussion at Help talk:Citation Style 1#Harmonizing CS1/CS2 into CS, and other cleanups for consistency?. Glades12 (talk) 04:58, 6 June 2020 (UTC)