Wikipedia talk:Citing sources/Archive 32

Recent changes were made to citations templates (such as {{citation}}, {{cite journal}}, {{cite web}}...). In addition to what was previously supported (bibcode, doi, jstor, isbn, ...), templates now support arXiv, ASIN, JFM, LCCN, MR, OL, OSTI, RFC, SSRN and Zbl. Before, you needed to place |id={{arxiv|0123.4567}} (or worse |url=http://arxiv.org/abs/0123.4567), now you can simply use |arxiv=0123.4567, likewise for |id={{JSTOR|0123456789}} and |url=http://www.jstor.org/stable/0123456789 → |jstor=0123456789.

The full list of supported identifiers is given here (with dummy values):

• {{cite journal |author=John Smith |year=2000 |title=How to Put Things into Other Things |journal=Journal of Foobar |volume=1 |issue=2 |pages=3–4 |arxiv=0123456789 |asin=0123456789 |bibcode=0123456789 |doi=0123456789 |jfm=0123456789 |jstor=0123456789 |lccn=0123456789 |isbn=0123456789 |issn=0123456789 |mr=0123456789 |oclc=0123456789 |ol=0123456789 |osti=0123456789 |rfc=0123456789 |pmc=0123456789 |pmid=0123456789 |ssrn=0123456789 |zbl=0123456789 |id={{para|id|____}} }}

Obviously not all citations needs all parameters, but this streamlines the most popular ones and gives both better metadata and better appearances when printed. Headbomb {talk / contribs / physics / books} 15:24, 17 March 2011 (UTC)

I post this earlier on several notice boards and Wikiproject talk pages, but I somehow missed this one. So here goes. Headbomb {talk / contribs / physics / books} 15:24, 17 March 2011 (UTC)

Is there a standard somewhere about whether to go through the references on an article and change all the citation author's first names to initials in the name of consistency? (I seem to recall a page about this but I can't track it down now.) I don't agree with this practice, but I would like to know the consensus first. Thanks.—RJH (talk) 17:10, 15 March 2011 (UTC)

We use the "rule of first editor": use the style established by the first editor to add the content in question; change only after discussion and consensus, to maintain consistency of subsequent edits or where the style is wildly against the norm. ---— Gadget850 (Ed) ^talk 17:25, 15 March 2011 (UTC)

See WP:CITEVAR. Consistency is good, although a talk-page discussion about whether full names provide more information to the reader (and is therefore an absolute good) might be warranted. WhatamIdoing (talk) 17:32, 15 March 2011 (UTC)

Found it. Per Wikipedia:Citing sources/example style: "A good guideline is to list author names as they are written in the original article/book, without further abbreviation. The APA guidelines recommend abbreviating first names to initial letters instead, but since Wikipedia has no shortage of space, you need not abbreviate names." I personally like to keep the full name because it helps with certain Wikipedia searches, as well as a few cases where the original link is lost.

The WP:CITEVAR is mainly about the general reference layout. I'm not clear it is as specific as whether to always use initials; that seems like overkill.—RJH (talk)

For context, this concerns a handful of references with first names amongst a sea of references with initials. Like ~15 in "Smith, John" style out of 150 (the rest being in "Smith, J." style). RJH argues that since a few are in "Smith, John" style, they should all be in "Smith, John" style (which will lead to both inconsistency as some of these references have authors that don't have full names listed in databases, and a great pain in the ass to find the first names of several hundred authors). I argue that since most are in "Smith, J.", then the odd ones should be streamlined to that format to conform to the article's established dominant style Headbomb {talk / contribs / physics / books} 18:19, 15 March 2011 (UTC)

No, I was arguing that the names should be in the same format as provided from the original source. I see no benefit to this streamlining process; information is being lost in the interest of conformity. There is also no specific policy in favor of this approach, and a pretty clear-cut statement against it in the example provided above.—RJH (talk) 18:27, 15 March 2011 (UTC)

The only information that is lost is irrelevant in the first place. Find me one professional Manual of Style which allows for inconstant citation styles. Chigaco doesn't. MLA doesn't. APA doesn't. AMA doesn't. CSE doesn't. Harvard doesn't. And consequently, WP:MOS (see "General principles: Consistency" and WP:CITEVAR in particular) doesn't. Headbomb {talk / contribs / physics / books} 18:43, 15 March 2011 (UTC)

Again, see the example. You're dragging in style guides that Wikipedia doesn't use; therefore your examples are irrelevant. Wikipedia has it's own standards that are governed in part by the principle that Wikipedia is not paper. This makes it a different medium than all others that have been governed by said guidelines.

Finally, just because you don't find the information useful doesn't mean that others don't. I have found the information useful in the past, particularly when performing Wikipedia searches, trying to recover a lost link or when I want to disambiguate an author for linking. (Smith, J. anybody?) The author information is also available via Google for everybody to use; you are diminishing that capability. This is unnecessary, unhelpful and needlessly destructive formatting.—RJH (talk) 19:18, 15 March 2011 (UTC)

RJHall, Wikipedia does use style guides; it uses whichever one chosen by the author who expands the article beyond a stub, or a style agreed to by a consensus demonstrated on the article's talk page. Jc3s5h (talk) 20:32, 15 March 2011 (UTC)

Ah, well that would be me then, since I provided most of the cites for these articles. No matter. I can understand that style guides would serve as a basis, but I also know that Wikipedia styles override said guides. Basically what I'm looking for here is a consensus on how to format citation authors. If it is to be dogmatic following of the style guides, then so be it. But I would like to see it written in stone so I don't have three months of work getting stomped on whenever somebody gets a style preference itch. :-) Thanks.—RJH (talk) 20:39, 15 March 2011 (UTC)

That is only your desire, not the article's dominant style. The vast majority of sources were in "Smith, J." style, so that was the established style. Headbomb {talk / contribs / physics / books} 01:57, 16 March 2011 (UTC)

If you look closely, you will see that although Wikipedia has a main Manual of style and several manuals of style for particular areas, none of them cover citations, except to refer to Citing sources. That, in turn, says to use the style chosen by the first editor to expand the article beyond a stub. There are families of templates that could be used for citations, such as {{Citation}} and {{Cite xxx}}, but those systems do not have a thorough manual of style describing how to use them, just some documentation pages with cursory descriptions of some of the parameters. In any case, something found in the documentation page of {{Cite book}} would have no bearing on an article that uses APA style.

A common problem is the author who starts an article uses some system of style, but never puts in a comment saying which style is being followed, so later editors have to try to guess from the format of the existing citations. Jc3s5h (talk) 00:10, 16 March 2011 (UTC)

Where possible, full names should be used; if the full name of an author isn't given, then you fill in what you can. I don't see why this is such an issue (and I certainly don't see why you would remove author info for the sake of conforming to initials from other sources.) Der Wohltemperierte Fuchs^(talk) 20:38, 15 March 2011 (UTC)

Not by a competent editor: E. A. Wallis Budge, the form he used himself - and, more important, by which reliable sources cite him - not Sir Ernest Alfred Thompson Wallis Budge, which will only leave readers wondering if this is some relation. Septentrionalis PMAnderson 23:05, 15 March 2011 (UTC)

WP:CITE allows any style. If an article is started with full names, that's fine, and if it's started with initials, that's fine. The style that's already present should be followed for subsequent references. In practice, in the presence of other citation information, initials rarely make a difference in locating a source. To get some confusion, you would need something like two different papers published in the same issue of the same journal on the same pages whose authors have the same last name :) So readers can locate the source with initials or with full names. — Carl (CBM · talk) 20:57, 15 March 2011 (UTC)

Yes, that is relatively true of modern papers. I have run into instances where there was ambiguity of some sources (such as older books or papers) and I needed to identify the actual author in order to establish credibility. Also useful when author wikilinks are desired, as mentioned previously.—RJH (talk) 22:11, 15 March 2011 (UTC)

Any style guide that objects to citing Peter Brown and W. W. Tarn on the same page, in the style in which their names will be recognized, is incompetent and should be discarded. Never let rules interfere with writing the encyclopedia. Septentrionalis PMAnderson 22:39, 15 March 2011 (UTC)

Yes, among the authors of reliable sources for British railway history we have David St John Thomas, who is almost never abbreviated to D. [St J.] Thomas (but is often written Thomas, David St John), and O. S. Nock, who is almost never expanded to Oswald [Stevens] Nock (but is often written Nock, O. S.). --Redrose64 (talk) 22:58, 15 March 2011 (UTC)

I would expect most journals would put all the references into their house style, which would either show initials for all authors or full names for all authors in the reference list. — Carl (CBM · talk) 00:43, 16 March 2011 (UTC)

That is indeed the case. I know of no journals, MoS, publisher, etc... which allowed to mix both styles. Headbomb {talk / contribs / physics / books} 01:57, 16 March 2011 (UTC)

Those two names are used together - in precisely those forms - by articles in The Classical Review and Greece & Rome, two of the leading journals in the field. Most journals mix initials and full names in their TOC, to say nothing of bibliographies. The true short form (single initial and surname) tends not to be mixed with other forms, but there are examples even of that (especially for Russian or French authors). Septentrionalis PMAnderson 02:11, 16 March 2011 (UTC)

For an example, see this page which tends to single initial and surname, but also uses full names, surname alone, and double initials. (N[icholas] Siouffi and most of the others cited wrote in French, and their papers generally have single initial and surname.) Septentrionalis PMAnderson 02:19, 16 March 2011 (UTC)

It looks like all the actual citations there use a single initial, citations to previously mentioned books use just the last name, and the full names are sentences, rather than in formal citations. — Carl (CBM · talk) 02:38, 16 March 2011 (UTC)

I was about to make the same observations as Carl, but he beat me to it. Headbomb {talk / contribs / physics / books} 02:58, 16 March 2011 (UTC)

• "A good guideline is to list author names as they are written in the original article/book, without further abbreviation. The APA guidelines recommend abbreviating first names to initial letters instead, but since Wikipedia has no shortage of space, you need not abbreviate names."—That is an invitation to make reference lists scrappy and inconsistent within an article. The guideline needs to be updated. First author's choice seems reasonable. I must say, spelling out first names is not the usual way in research publications, but if editors are keen to do this in an article, I suppose there's no pressing reason not to. I guess the middle name would then be represented by an initial, if anything? Tony (talk) 07:01, 16 March 2011 (UTC)

It's a pretty bad guideline actually. If you're citing a paper which lists its authors as "Smith, JC; Jones, T; Morgan, PR", then another one as "Smith, Jim C; Jones, T; Morgan, Paul, R", then other one as "J. C. Smith, T. Jones, P. R. Morgan" as they are themselves cited, you have a horror on your hand. Likewise if you use "Smith, JC; Jones, T; Morgan, PR" half the time, and "Smith, Jim Charles; Jones, Tim; Morgan, Paul Ruben" the other half. Following the original author's style is a bad idea if that style is mean inconsistency. Headbomb {talk / contribs / physics / books} 08:10, 16 March 2011 (UTC)

Re Tony's "spelling out first names is not the usual way in research publications", there's no consistency between research publications in zoology and paleontology. Most journal articles give the author(s) name(s) as "first-name middle-initial surname" but a few journals give only first initial, e.g. The Verdun Syndrome: simultaneous origin of protective armour and infaunal shelters at the Precambrian–Cambrian transition gives the author's name as "J. Dzik" under the title on the first page, and Journal of the Geological Society gives only initials, e.g. Small shelly fossils of late Precambrian and early Cambrian age: a review of recent work. On the hand, all journal articles that I can remember give only initials in lists of cited articles at the bottom of the current article, and PNAS and Royal Society pages use initials for the citation details of the current articles as well as showing the full name under the title.

I've just found another variant: initial, full middle name, full surname at Tardigrades survive exposure to space in low Earth orbit --Philcha (talk) 12:00, 16 March 2011 (UTC)

Sometimes you can't find a full first name, e.g. for Small shelly fossils of late Precambrian and early Cambrian age: a review of recent work I found on Google the first name of the 1st author but not of the 2nd author.

In books there often multiple authors, e.g. "Invertebrate Zoology: ..." by E.E. Ruppert, R.S. Fox and R.D. Barnes. In such cases full first names (which I know, I have the book) would just clutter the "References" section of the WP articles. For reductio ad absurdum, I remember seeing a journal articles with about 20 authors.

But we might want to use full firstnames if e.g. there are 2 "J. Smith"s.

I think there's no feasible general rule, and hopefully common sense reigns. --Philcha (talk) 11:03, 16 March 2011 (UTC)

There is certainly a need for consistent layout of the information, but I remain unconvinced of a need to enforce the use of initials for first names when the source provides the full name. Dogmatically enforcing the use of initials for first names removes value from an article and adds nothing. Now if you wanted to say that first names should only be allowed when the list of authors is brief (say 1-3 authors), I could see that as useful. But for, say, a single-author old book, news story or (especially) a web page, the full name adds value and should be kept IMO.—RJH (talk) 14:15, 16 March 2011 (UTC)

I also argue for consistency, but there are any number of authors where the style of their name is a direct connection to their works. For example, J. R. R. Tolkien would never use the full first name, but Edgar Rice Burroughs always uses the full first and middle names. I have never seen Arthur C. Clarke or John W. Campbell without the middle initial, but Isaac Asimov and Alfred Bester never used middle initials. There there is L. Sprague de Camp who never used a full first name. A few more off the top of my head using initials only: E. E. Smith, E. E. Cummings, F. M. Busby, T. S. Eliot. Authors using full names: Roger MacBride Allen, Lois McMaster Bujold. ---— Gadget850 (Ed) ^talk 15:27, 17 March 2011 (UTC)

If I am interpreting Headbomb's viewpoint correctly, then he is saying, in effect, that Wikipedia's WP:CONSISTENCY policy means that all author names should be listed the same way in the references block. In other words, if at least one of the citations only has first initials available, then all citations must use initials for the first name. It is an all or nothing policy. Hence, your list would be: Allen, R. M.; Bujold, L. M.; de Camp, L. S.; Smith, E. E., and so forth.—RJH (talk) 16:25, 17 March 2011 (UTC)

Haven't read all the previous stuff. I am in favor of a standard (names or initials, don't care). Academice journals differ in if they use names or initials (and even if they put periods on the initials, run them together, etc.) But they follow the same format WITHIN A JOURNAL. So let's pick a style for the journal of Wikipedia and use it. And don't tell me we do it by the article...that is a clusterfuck and we know it doesn't work....the articles endup a mishmash of styles inside the articles and it requires people to learn 10 different styles. Let's standardize.

My inclination would be to go for full names for the same rationale that we don't generally abbreviate journal names and that we do generally give article names instead of just pages. It's because we are not paper. And a fuller citation better allows the reader to decide if he should try to get the cite, and it really helps with all the verification and such. Of course, sometimes we won't have a full name, but no biggie, use it when we have it. But again, if we can standardize, I am totally fine with initials too.

TCO (talk) 06:36, 19 March 2011 (UTC)

Journal articles tend to cite other journal articles, so using a single standard is easier for them. I don't think we can use an identical standard across Wikipedia because the various academic disciplines tend to use their own name convention. I think we just need to leave some room for exceptions, minor but consistent variations, and common sense.—RJH (talk) 18:12, 19 March 2011 (UTC)

Per the prior discussion, what I'd like to propose is to insert the following compromise recommendation into the Style variation section of the style guide:

"For references with long lists of authors, for brevity it is recommended that the full last name and the first and middle initials be retained. In references where a single author is being listed and the last name may be ambiguous, the full name of the author should be used. When a publication has a single author and is out of print (or likely to go out of print, such as a web page or news article), the full name should be used. In all other situations, use your best judgment."

Alternatively, we could just list cases where we would want to retain the full author name. Thanks.—RJH (talk) 14:44, 17 March 2011 (UTC)

Absolutely not. Editors should be free to choose whatever style they feel like, as long as it is consistent throughout the article. The full names work on some article, but certainly not on most. Headbomb {talk / contribs / physics / books} 14:57, 17 March 2011 (UTC)

No. It's not what "we" want: It's what they want. Editors are free to use any citation style, and this means we give them the freedom to go look up a hundred authors' first names if they want, and the freedom not to, if they don't want. There are both technical (PubMed lists only initials) and cultural (using initials is socially normative in India) reasons to let editors do whatever they want on this point. WhatamIdoing (talk) 15:04, 17 March 2011 (UTC)

Ah no, we still disagree. Where you say consistent "style", I think consistent "layout". The description of the Style variation section does not get down to the fine detail of how names should be presented and the style guide example page quite clearly shows that the use of mixed names formats is acceptable. Ergo, editors should be free to apply a consistent naming convention, but not necessarily a blanket consistency as you would prefer.—RJH (talk) 15:06, 17 March 2011 (UTC)

WP:MOS, and particularly WP:CONSISTENCY apply here, so no you shouldd not mix citation styles. Headbomb {talk / contribs / physics / books} 15:08, 17 March 2011 (UTC)

Right then, per the example, "A good guideline is to list author names as they are written in the original article/book, without further abbreviation." This is a consistent style and a consistent approach. We aren't disagreeing here; we're just looking at the meaning differently.—RJH (talk) 15:22, 17 March 2011 (UTC)

It is neither consistent, nor a good guideline to be applied on well-developed articles. Headbomb {talk / contribs / physics / books} 17:26, 17 March 2011 (UTC)

Yes we disagree. In this case, the policy in the second paragraph of "Style variation" applies. Hence, we fall back to using full names on the Supernova article.—RJH (talk) 17:30, 17 March 2011 (UTC)

We most certainly do not fall back on using an inconsistent style based on a minority of citations. That goes directly against WP:CITEVAR.Headbomb {talk / contribs / physics / books} 17:44, 17 March 2011 (UTC)

I assume this is the royal "we", and your opinions certainly don't appear in the link you cited. Very well. "As with issues of spelling differences, if there is disagreement about which style is best, defer to the style used by the first major contributor." That would be myself and WolfmanSF. I pretty much got this article cited up to FA standard, so probably me then.

I can certainly scratch your itch about consistency by consistently inserting full names, that is assuming you aren't going to engage in more edit warring.—RJH (talk) 18:46, 17 March 2011 (UTC)

It is the same "we" you used. As far the style that was used by the article, "Smith, J." was the dominant style. You make claim you wanted to use another, but that was the dominant style used, so that's the one that will be used for consistency. Headbomb {talk / contribs / physics / books} 18:59, 17 March 2011 (UTC)

The "we" I used just above is the "you and I" we, as in you and I disagree. No, the dominant style was to use the name provided by the source, as per the example page. In particular, I generally used full names for books and web pages (where possible) and initials for journals; different types of citations and thus different styles.—RJH (talk) 19:01, 17 March 2011 (UTC)

Mixing "Smith, J.C." "Smith JC" "J. C. Smith" "John C. Smith" based on the styles of the original citation is not a recognized citation style anymore than writing "Jim was found dead on March 9, 2009. Two days after, on 2010-03-10, Jim told Bob that he would bring him a toaster by 29 March 2010, but possibly earlier (18-Mar-2010)" based on what the sources write is a recognized date style. Headbomb {talk / contribs / physics / books} 19:10, 17 March 2011 (UTC)

Okay, I think I see where there is a problem in our discussion. You and I may be interpreting the example differently; to me it means to use each of the individual words of the names as they appear in the source: Smith, John Q. rather than Smith, J. Q., when the source says John Q. Smith. For me this follows from the statement that, "Wikipedia has no shortage of space, you need not abbreviate names." Yes, of course we should use a consistent layout. I'm not disputing that. My only concern is with the format of the first (or middle name).—RJH (talk) 19:21, 17 March 2011 (UTC)

I'm not sure that you understand my position. If the editors at the article make up a citation style that says "Use whatever name is on the source", then that's the standard they should consistently follow throughout the article. If they editors at the article make up a citation style that says "Always use full names", then that's the standard they should consistently follow throughout. If they editors at the article make up a citation style that says "Always use initials for first and middle names, even if we know the full name", then that's the standard they should consistently follow throughout.

They can make up any style they want. Once they—not "we here", but "they there"—have decided what style they want, then they should follow it consistently throughout the article. "We here" do not tell them what that style should be. We only tell them to make up their minds and stick with whatever style they freely chose (or invented). WhatamIdoing (talk) 17:22, 18 March 2011 (UTC)

Right on. Now, anything left that is specific to Supernova should be taken to Talk:Supernova, then perhaps my watchlist will be less cluttered. --Redrose64 (talk) 17:49, 18 March 2011 (UTC)

I would have taken it there, but this has happened on at least one other FA article and possibly many. Hence, why I requested to discuss it here.—RJH (talk)

I agree. I haven't disagreed with you (as WhatamIdoing). Nor should an editor come along and force a different consistency model on the editors without first gaining consensus.

However, there are many ways to be consistent. If you follow the model that says use full first names when available, or initials when not, then this is a form of algorithmic consistency. If you are going to rigidly enforce a brand of monolithic consistency, then you must do that consistently. Thus, et al. should always be used consistently according to some criteria (whenever there are 3+ authors); capitalization of the reference titles should always be performed consistently; the presentation of dates and access dates must also be monolithically consistent with each other, and so forth. Personally, I think that approach is pedantic beyond all reason, but there you go.—RJH (talk) 17:52, 18 March 2011 (UTC)

"Algorythmic consistency" is utter nonsense. Jim was found dead on March 9, 2009. Two days after, on 2010-03-10, Jim told Bob that he would bring him a toaster by 29 March 2010, but possibly earlier (18-Mar-2010) is not a consistent style because whoever wrote it "consistently wrote the dates as he found them in the sources". It's an utterly and completely inconsistent "style" (and I use the term loosely here) because none of the dates are in the same format. It is exactly the same situation if an article has refers to "John Smith", "Jones, C.", "Daniel K", "O'Neil, F" and "M BOUCHER" because the names are "algorithmically" taken from five different sources, each with a different style. Consistency is picking one style, and sticking with it throughout the article. Headbomb {talk / contribs / physics / books} 21:37, 18 March 2011 (UTC)

Yet more reductio ad absurdum, but only fair since I did the same. :-) At any rate, I remain unconvinced by your arguments. See WhatamIdoing's comments above. Consistency is whatever principles, course or form that the editor's have agreed upon, or whatever principles, course or form was first established. If the principle was to retain first names where they appear, then that is consistent behavior.—RJH (talk) 21:55, 18 March 2011 (UTC)

• Oppose. We ought not recommend specific style (initials, full names, what comes first, punctuation, etc.). Editors choose the style for an article the way they decide. Different fields cite differently. So the editors decide how they cite consistently in the article. —  HELLKNOWZ  ▎TALK 22:12, 18 March 2011 (UTC)

• Oppose. This is playing "my style is better than your style." Fifelfoo (talk) 23:38, 18 March 2011 (UTC)
  • Fair enough, I withdraw my suggestion. I suppose I was attempting to reach a gentlemanly compromise with HeadBomb's position, which appears impossible in any event. Thank you for the feedback.—RJH (talk) 18:01, 19 March 2011 (UTC)

Open Library has an option to get a full Wikipedia citation for any book in it's catalogue, for example http://openlibrary.org/books/OL13538404M/Der_Steppenwolf . I think we should mention that somewhere, because it's a really handy and simple way to get citations for books. Any suggestions? Sadads (talk) 14:12, 13 March 2011 (UTC)

What do you mean "has the option to get a full Wikipedia citation" ? Headbomb {talk / contribs / physics / books} 14:16, 15 March 2011 (UTC)

If you look down the page you'll see the text "Download catalog record: RDF / JSON / OPDS | Wikipedia citation". Clicking the last pops up a box headed "Wikipedia citation Copy and paste this code into your Wikipedia page. Need help?", containing

{{Citation
    |publisher = Suhrkamp
    |publication-place = Frankfurt am Main
    |title = Steppenwolf
    |url = http://openlibrary.org/books/OL13538404M/Der_Steppenwolf
    |author = Hermann Hesse
    |edition = Der Steppenwolf
    |publication-date = 1960
    |oclc = 6584578
    }}

which will need tweaking to get it to fit in with common citation styles (for example, we'd put the author as "Hesse, Hermann"), and also for Harvard ref linking to work. Also, {{cite book}} is far more commonly used than {{citation}}, and cite book has the advantage that it handles the |trans_title= parameter, which citation doesn't (hence the misuse of |edition=). --Redrose64 (talk) 14:40, 15 March 2011 (UTC)

I am on the mailing list if you want any changes to the format, I can make a request. I would prefer them using the "|first =" and "last =" variables to fix the name thing, but I am not sure how flexible their output is right now. I think this could be very useful for people who want to add citations to books, even if there are small problems with the format, at least it includes most of the more important information that will be needed for future editors to clean up the ref. Any input anyone has would be great, and I can compose an e-mail with a link to the discussion here once we know what we want changed, Sadads (talk) 18:23, 15 March 2011 (UTC)

Hmmmm... I see. The citation should most likely use |ol=13538404M rather than the URL, since that will lead to better appearance and tell people the link will take them to OpenLibrary. It should also make better use of the parameters (for this book, the following would be optimal), and give the option of multiline

{{Citation
 |author=Hermann Hesse                          <!-- Or possibly |last=Hesse |first=Hermann       -->
 |year=1960                                     <!-- Or possibly |date=1960 to simplify compatibility
 |title=Der Steppenwolf                              with full dates such as 18 February 1974     -->
 |series=His Gesammelte Werke in Einzelausgaben
 |language=German                               <!-- Which should be omitted if it's English      -->
 |publisher=Suhrkamp
 |location=Frankfurt am Main
 |oclc=6584578
 |ol=13538404M
}}

or single line

{{Citation |author=Hermann Hesse |year=1960 |title=Der Steppenwolf |series=His Gesammelte Werke in Einzelausgaben |language=German |publisher=Suhrkamp |location=Frankfurt am Main |oclc=6584578 |ol=13538404M}}

to produce

• Hermann Hesse (1960), Der Steppenwolf, His Gesammelte Werke in Einzelausgaben, Frankfurt am Main: Suhrkamp, OCLC 6584578, OL 13538404M

Headbomb {talk / contribs / physics / books} 07:38, 17 March 2011 (UTC)

Re the comment "Or possibly |date=1960 to simplify compatibility with full dates such as 18 February 1974" - please don't use |date= for a pure year, 60% of the time it's misinterpreted and Harvard reference linking fails; as it happens, 1960 falls within the 40% where |date=1960 does work. Further info at Template:Harv#Wikilink to citation does not work item 2.1.1.4 and the linked note.

In connection with Harvard linking, this works best if separate |last=|first= are used, since otherwise the Harvard ref will need to contain the full author name. Compare (Hesse 1960, p. 123) harv error: no target: CITEREFHesse1960 (help), which doesn't link to the above citation, with (Hermann Hesse 1960, p. 123), which does. --Redrose64 (talk) 14:18, 17 March 2011 (UTC)

That stuff can be fixed after the citation is inserted. I'm suggesting |date= and |author= because it might not be possible to ensure data integrity if |last= |first= and |year= are used. If |last= |first= (and |last2= |first2= ...) can't be filled properly, then |author= should be used. If |year= can't be filled properly, |date= should be used (AWB and Bots will convert |date= into |year= when appropriate anyway). Headbomb {talk / contribs / physics / books} 15:01, 17 March 2011 (UTC)

Sounds good, I will have them modify the output then into the more useful/common parameters. Thanks for the feedback everyone, Sadads (talk) 17:21, 21 March 2011 (UTC)

How does one make it clear that a given ref is applicable to the entire content of a section, or a whole list, table, paragraph, etc? Roger (talk) 19:51, 22 March 2011 (UTC)

You can either just add it at the end of the paragraph, and let readers draw the obvious conclusion. Or you can add to the footnote something like: "For the source supporting the preceding paragraph, see Smith, John, etc." Or you can list the different points in the footnote: "for the date of birth, the day of the party, and the reason for the arrest, see Smith, John, etc." SlimVirgin ^{TALK|CONTRIBS} 20:11, 22 March 2011 (UTC)

The problem occurs when the paragraph is modified or expanded at a later date. The citation is now an insufficient source, except possibly for the final sentence. I've also experienced situations where a paragraph will be properly tagged with a single cite, but then some editor comes along and adds a citation needed in the middle of the paragraph, not knowing that the cite applies to all.—RJH (talk) 21:12, 22 March 2011 (UTC)

I think a problem is that some editors just don't get the concept of supporting statements with citations. Others don't understand the way citations are generally used in academic writing. If editors don't understand these basic concepts, what is the chance they would take the time to decipher some newly-invented Wikipedia-only citation convention that specifies exactly what statement a citation applies to? My opinion: no chance at all. The only way to fix damage from editors who don't understand citations is for editors who do understand to repair the damage. Jc3s5h (talk) 21:47, 22 March 2011 (UTC)

Yes. I was contemplating some form of hidden XML-style tags to designate sections cited by sources. But that would be an implementation and maintenance nightmare. I guess the citation stability and reliability of Wikipedia will continue to depend on editor-custodians who have in-depth knowledgeable about the pages that they watch.—RJH (talk) 21:56, 22 March 2011 (UTC)

I've occasionally left <!-- hidden comments --> in the text if I think future editors will be confused, or comments on the talk page if I think the current editors will be confused. For a table, I usually think it best to put the citation in a caption. For a list, if it leads in with a phrase ("The following are foo:") then it's usually good enough to list the citation after that phrase.

If it's just one paragraph (clearly about a single subject, without direct quotations, contentious matter about BLPs, etc.), then one copy of the citation, usually placed at the end, is enough. In this instance, however, some editors place a citation after both the first and the last sentence in the paragraph. WhatamIdoing (talk) 23:58, 22 March 2011 (UTC)

First and last seems very sensible. I, and many others it seems, just throw in a reference every three or four sentences (when a large section is all from one source) which seems to make things clear.  ⊂| Mr.choppers |⊃  (talk) 02:03, 24 March 2011 (UTC)

The explanatory text at Template:Cite web states that it is to be used to cite online sources, but that Cite news can also be used when citing a news source. This seems to clearly indicate that if your source is an 'online newspaper' you can use either template. However, see this bot revision has changed the reference from a cite web to a cite news. Is there any consensus for this? Eldumpo (talk) 12:04, 9 March 2011 (UTC)

Check the archives fro previous discussion. You could use {{cite web}} for a book, a journal or a news source, but the other templates have parameters that are more appropriate. Bottom line: use {{cite web}} only for a web page when one of the other templates doesn't fit. ---— Gadget850 (Ed) ^talk 12:49, 9 March 2011 (UTC)

Thanks for your response. I did some checking of the archives but could not see any similar discussions. Can you provide a link, or at least an approximate date? However, I note your comment that other templates are more appropriate and that cite web is more of a last resort if others don't fit. If this is regarded as the consensus then the introduction to the template should have appropriate rewording to indicate this. Hopefully there will be some further responses on here as to whether this is how Cite Web is supposed to be used. Eldumpo (talk) 11:23, 10 March 2011 (UTC)

An additional consideration that should be mentioned is cite web and news output some parameters differently, meaning that consistent citation schemes can be difficult; I've had editors just use {{cite web}} entirely for the reason that the vast majority were web sources. Der Wohltemperierte Fuchs^(talk) 19:47, 13 March 2011 (UTC)

Although I understand the standardization argument and advocate standardization whenever I can I also believe that using Cite news for news sources is more appropriate than a simple cite web because of the extra fields that the Cite news template provides. I would like to see the style of both templates be the same though that is a different discussion. --Kumioko (talk) 19:54, 14 March 2011 (UTC)

I was previously asked by another editor (forget who, it must have been more than a year ago) for my bot task to convert {{cite web}} to {{cite news}} when adding cite news instances for the same newspapers to aid overall article consistency. Rjwilmsi 20:03, 14 March 2011 (UTC)

OK, there seems to be consensus that news should be used in preference to web. I therefore propose the following text should be included as the intro to Template:cite web.

"This template is used to cite online sources in Wikipedia articles. However, it is preferred that {{Cite news}}, {{Cite book}} and {{Cite journal}} are used instead for those citations that are online versions of newspapers, books or journals. For general information about citations in Wikipedia articles, see Wikipedia:Cite sources. A general discussion of the use of templates for adding citations to Wikipedia articles is available at Wikipedia:Citation templates."

Any comments on this text, and also, is it OK to just go in and edit the template (given it is in use on so many pages), or does it need to be done by an admin? Eldumpo (talk) 22:59, 15 March 2011 (UTC)

Just do it. There is consensus. -- Magioladitis (talk) 23:22, 16 March 2011 (UTC)

I'm going to point out here that it's not at all how many projects go about it, particularly WP:VG from my area of expertise. @Kumioko, what extra fields are necessary than what cite web provides? Work, publisher, page, accessdate, date, author, url, title are all in cite web. Der Wohltemperierte Fuchs^(talk) 23:31, 16 March 2011 (UTC)

Per your request Eldumpo here is a table comparing the parameter differences between the three citation types above. As you can see, although all three have fields in common each has fields the others do not. The wording looks good to me though.

--Kumioko (talk) 01:18, 17 March 2011 (UTC)

Is the above based upon what the documentation lists, or what the templates actually recognise? --Redrose64 (talk) 09:22, 17 March 2011 (UTC)

Thats a very good point. I went by what the documentation says but its possible that the documentation isn't accurate and needs to be updated. --Kumioko (talk) 20:41, 19 March 2011 (UTC)

Der Wohltemperierte Fuchs asks "what extra fields are necessary than what cite web provides? Work, publisher, page, accessdate, date, author, url, title are all in cite web." In "cite news" you have "newspaper". This comes out the same as "work" but it is clearer what it means. Why it is important to use "cite news" and not "cite web" is that the title of the newspaper or magazine etc. then comes out in italics, which is the required form under WP:ITALIC. (This is also why it is important to put the name of the source in "newspaper" or "work" and not in "publisher", which in the context of "cite news" means something different and is not normally needed. I seem to spend half my life correcting references where editors have not understood this. I wish it could be made clearer somehow.) -- Alarics (talk) 11:12, 29 March 2011 (UTC)

All of these cite templates use {{Citation/core}}, just making appropriate aliases to map their named parameters to the ones used by Citation/core. You can look at the source for e.g. {{Cite web}} and see what parameters it actually supports. For example, Cite web supports 'authorlink2' to 'authorlink9', regardless of what the documentation says. The only reason for these different templates is that it should make it easier for editors to pick the appropriate information to supply to each named field. I would agree that if they fail to do that, then changes should be made. --RexxS (talk) 16:21, 29 March 2011 (UTC)

Most editors know nothing of citation/core, and should not need to. The "visible facts on the ground" are that if you use "cite web" for a news citation, it comes out looking different from if you use "cite news", and furthermore it comes out looking wrong, if you try to use the "newspaper" parameter and/or if you put the name of the newspaper under "publisher" instead of under "work". (This is because the name of the newspaper should come out in italics.) Which brings us back to the point of the original question in this thread, which was why can we not spell out that "cite news" SHOULD (not just may) be used for a news reference (whether online or not), not "cite web"? -- Alarics (talk) 19:33, 29 March 2011 (UTC)

It's certainly a good idea to recommend using the template which is most likely to yield good results for the inexperienced. From that view, {cite news} ought to be preferred to {cite web}. The only problem is the pedants who will then make a career out of changing hundreds of perfectly well-formatted {cite web} templates to {cite news} because you SHOULD use it for a news reference – which does sometimes seem to annoy the original authors. If you can find a form of wording that met the recommendation, while avoiding the the problems identified, I'd be keen to support it. --RexxS (talk) 22:29, 29 March 2011 (UTC)

"The only problem is the pedants who will then make a career out of changing hundreds of perfectly well-formatted {cite web} templates to {cite news}" -- well, I am trying not to make a career out of it, but I do often find myself making that change, not because of pedantry but for the sake of consistency: one usually finds the references in any given article are a complete dog's breakfast of inconsistent, and very often wrong, citation styles. Most editors seem to understand little about referencing and/or don't care. So I don't think your "problem" is much of a problem in reality. -- Alarics (talk) 12:38, 30 March 2011 (UTC)

I think you're on very safe ground changing a citation when it brings the visual result into consistency with other references. I think you'll find yourself in conflict if you work on the premise that some citation styles are "wrong" – particularly if you tell other editors that their choice of {cite web} for an online news source is "wrong" if it is producing correctly formatted output. I don't think you'll persuade other editors on this page to prohibit the use of {cite web} for an online news source. --RexxS (talk) 13:13, 30 March 2011 (UTC)

Well, I always thought that already was the rule, but cannot now remember where I saw it. I have been proceeding on that basis for a long time and nobody has ever complained. -- Alarics (talk) 08:52, 31 March 2011 (UTC)

I undid an edit that changed this guideline to say that the use of "named" references is mandatory when the same footnote content is used more than once. My reasons:

• It seems to go against CITEVAR. Historically at least we haven't required the use of named references, and we usually leave that sort of choice up to individual articles.
• AWB intentionally does not add names to "duplicate" references in an when there are none already in an article, because the editors of the article may have chosen not to use them.

If there is some general agreement that using named references is mandatory, then by all means we should clarify it in this page or at WP:FOOTNOTES. But it seems strange to me to mandate that.

Of course, if we do mandate it, a bot will take care of implementing named references for all duplicate footnotes in all articles; the change to this guideline was by a bot operator who has proposed running that task. So we have to keep in mind whether that's desirable. Other people use footnotes more than I do, so I'd value some other opinions. — Carl (CBM · talk) 01:15, 19 March 2011 (UTC)

I think this is nothing to mandate but to recommend. Mandating such things is imho a case of WP:CREEP and WP:Bureaucracy.--Kmhkmh (talk) 01:58, 19 March 2011 (UTC)

A small number of editors seem to really, really hate named ref tags. I see no good reason to impose my personally preferred style on them. WhatamIdoing (talk) 02:43, 19 March 2011 (UTC)

I really really hate them, both as an editor and as a reader. Rjensen (talk) 03:12, 19 March 2011 (UTC)

Can I ask why (as a reader)? I assume because they generate a clutter of "^{a b c d e f g h} Smith. J....."? —  HELLKNOWZ  ▎TALK 08:27, 19 March 2011 (UTC)

It should be noted that there is at least one bot (citation bot 1 (talk · contribs) springs to mind) which is naming and thus merging duplicate refs, as with this edit. --Redrose64 (talk) 15:07, 19 March 2011 (UTC)

The one aspect of named ref tags I don't particularly like is the arbitrary nature of the name value. The values aren't always very communicative and sometimes an article ends up with duplicate names. Plus I have seen some obscenely long lists of backlinks. Otherwise... they do the job, so that's enough for me.—RJH (talk) 22:32, 24 March 2011 (UTC)

I agree. I hate working with something like <ref name="CROG199"/>. You can't tell what it is. And it's utterly unintelligible to new editors. ---- CharlesGillingham (talk) 06:21, 25 March 2011 (UTC)

I really hate named refs too. I'm in a dispute over another editor about them now. I think it would be good to include something stating that they are optional. It seems like more and more editors are pushing them, but I don't really get their appeal. If you click on a footnote, it takes you down to the citation, but then you have to choose from a, b, c... to click to get back to your spot. It's much easier to jump back and forth with single citations. Also, I like eyeballing citations to see which citations are used in which sections. You can't do this in articles with named references. Can some explain their appeal besides fewer numbered citations?--Bkwillwm (talk) 05:28, 25 March 2011 (UTC)

well one (obvious) advantage of them is that they make editing the source text of an article easier - in particular the source text gets easier to read, since they allow rather short inline citations (compared to a full citation at least). The latter make the source text often hard to read.--Kmhkmh (talk) 10:43, 25 March 2011 (UTC)

I think another appeal is you can see at a glance how often an editor has relied on one source, and whether it's over-reliance. SlimVirgin ^{TALK|CONTRIBS} 19:48, 25 March 2011 (UTC)

I still, personally, don't think named refs help that much. Regarding Kmhkmh's points, yes, they are better than full inline citations that clutter the source text, but an editor can simply use short citations and have a separate references section at the bottom of the article. Using a named ref lets you use <ref name="..."/> instead of <ref>...</ref>. Basically, you don't have to close the tag with ref names, but you do have to manage a standardized list of ref names, which can be confusing especially to editors new to an article. Regarding SlimVirgin's points... Really? This is actually one of the main reasons I dislike named refs. In regular, short inline citation articles, I can go to a section; see that it uses ref 40-48; jump to the references section, and then look at whether it is overly dependent on that source. In ref named articles, a section might use citation numbers from all over the article: 56, 3, 6, 22, 33. These references are scattered all over the references section, and I have to click through each one or write down a list of randomly ordered numbers to look up. True, ref names makes it easy to see if one page is used too many times, but I think usually the issue is whether one source is used too much, and it's easier to see that with standard short inline citations.--Bkwillwm (talk) 18:09, 26 March 2011 (UTC)

This discussion is not about whether short inline refs are better or worse than full inline refs - it is about whether or not inline refs (of whatever length) should be named or not. That is, should <ref name=Smith2001p123>Smith (2001) p. 123</ref> be preferred to <ref>Smith (2001) p. 123</ref>. --Redrose64 (talk) 15:49, 25 March 2011 (UTC)

You somewhat missed the point above. The poster above asked regarding the advantages of named references and why many (?) editors prefer them. And one reason for that is that for people preferring full references they are much shorter and even for people using short inline refs they are still somewhat shorter (<ref name="Smith123"/> versus <ref>Smith (2001) p. 123 </ref>).--Kmhkmh (talk) 18:46, 25 March 2011 (UTC)

What would you use as an alternative, given a need to uniquely identify specific references?—RJH (talk) 17:52, 25 March 2011 (UTC)

Me? It depends on the existing style, if there is one. If the article already uses inline <ref>...</ref> tags, I name them but only where necessary to consolidate duplicate refs - if an inline ref supports only one paragraph, sentence or phrase, I leave it un-named. For new articles, I use {{sfn}} instead. This names all inline refs, whether necessary or not: but the crucial thing is that it's done invisibly, see Wolf's Castle Halt railway station in particular the Parker & Morris 2008 refs. --Redrose64 (talk) 18:32, 25 March 2011 (UTC)

Yes I agree, the {{sfn}} is pretty nice. Perhaps we need something equivalent for people who use full inline cites? (I.e. a template that requires just enough fields filled out to uniquely correlate it with a full cite.)—RJH (talk) 18:44, 25 March 2011 (UTC)

A user (User:Fleetham) who likes to provide multiple references for every sentence when they edit, has picked up on the possibility of bundling references. However, since they provide so very many references and often the same one repeatedly, the same reference is often quoted several times in the reference. For an illustration, the reference section of the BYD Auto article currently shows a Wall Street Journal article ("Beijing Halts Construction of BYD Auto Plant") FOUR times and several other cites are listed twice or more.

To me, the obvious solution is to provide less references, but the other user is adamant about providing citations for everything. Is it best to bundle and cite the same reference over and over again, or are multiple superscripts preferred?  ⊂| Mr.choppers |⊃  (talk) 16:13, 23 March 2011 (UTC)

I don't know my first reaction is : live and let live, we don't have a mandated citation style and bugging the primary author of article regarding his particular citation style is almost always a bad move.--Kmhkmh (talk) 16:30, 23 March 2011 (UTC)

It's not just one page, the editor in question is targeting a large number of pages (and also changes my edits to match his preferred style, which is why I was hoping for some outside opinions).  ⊂| Mr.choppers |⊃  (talk) 16:41, 23 March 2011 (UTC)

If he is systemically changing citations in articles where he is not a primary editor then that is a different issue. Generally speaking annotated bundled citation are the most accurate form for a citation and they clutter less then individual citations. The drawback however is the switching or replacing of sources requires more work later on. The preferences differ author from to author and as I said above there is no mandated style and in doubt or in conflict people should follow the citation style that is already established in an article.--Kmhkmh (talk) 17:14, 23 March 2011 (UTC)

Mr. choppers, you have my sympathies. If he is not the primary editor for most of those articles and the citation style is already in good shape, then he may be heading toward a pattern of disruptive editing. You might check WP:DDE for suggestions on how to proceed.—RJH (talk) 20:19, 23 March 2011 (UTC)

I'm trying to avoid that, as there has been some seemingly fruitful conversation recently. The pages in question are of very narrow interest (Share taxi, Brilliance Auto, and most may be downright boring to most) so it's hard to convince anyone else to have a look see.  ⊂| Mr.choppers |⊃  (talk) 02:00, 24 March 2011 (UTC)

Understood. Thanks.—RJH (talk) 22:25, 24 March 2011 (UTC)

How is it done? Kindle does not display page numbers, which is a huge pain in the ass for any student in class who has to refer to a certain page or reference it for a research paper. Instead they have these worthless "location numbers" which as far as I know are not accepted outside of Kindle in any standard referencing. Are location numbers acceptable in a citation on Wikipedia? I would not be surprised if the answer is a big resounding no. And if not, do you cite Kindle like you would a webpage or like a theatrical play? i.e. "Chapter 5, Section 2, Paragraph 8" etc. Thanks to anyone who knows anything about this subject and is willing to answer my questions. Cheers.--Pericles of Athens^Talk 20:45, 24 March 2011 (UTC)

Just to offer an opinion. If there is print version of a book, if possible try to cite that one. Otherwise use whatever description works to narrow down the content in question for that "Chapter 5, Section 2, Paragraph 8" ais fine, depending on what you want to source even a reference by chapter (only) might be sufficient. The location number should be avoided or rather just given as additional info, since it can only be used by people reading the same ebook version, which makes it unconvenient for other users. However I wouldn't go so far as saying the location number is unacceptable, since the overal goal is simply to provide a precise enough citation for verification and the location number provides that. Strictly speaking even page numbers might be only correct for a particular edition, so the principal problem is the same there.--Kmhkmh (talk) 21:12, 24 March 2011 (UTC)

See previous threads at Wikipedia talk:Citing sources/Archive 29#Citing sources from ebooks and Template talk:Cite book#eBooks and page numbers. --Redrose64 (talk) 21:24, 24 March 2011 (UTC)

Thanks, guys.--Pericles of Athens^Talk 22:31, 24 March 2011 (UTC)

Just for information: Amazon adding 'real' page numbers to Kindle (February 7, 2011) -- PBS (talk) 22:49, 26 March 2011 (UTC)

I have a source coming from the United States Census Bureau, but it is convenience linked on Ancestry.com. I have checked with WP:RSN and it (Ancestry.com) is a Reliable Source (since it is actually coming from the Census Bureau). My question is, who do I source in the source template? Ancestry.com, the US Census Bureau, or both? - Neutralhomer • Talk • Coor. Online Amb'dor • 22:13, 25 March 2011 (UTC)

WP:SAYWHEREYOUGOTIT. --Redrose64 (talk) 17:33, 26 March 2011 (UTC)

Done, thank you. :) - Neutralhomer • Talk • Coor. Online Amb'dor • 17:39, 26 March 2011 (UTC)

Is there any guidance/best practice for the use of the accessdate parameter on the citation templates. If you click on an existing reference which already has a 'retrieved on...' entry, should you automatically update the date to the day you checked it, or should you only update the access date if you make other changes to the citation (title, author etc)? Eldumpo (talk) 10:05, 29 March 2011 (UTC)

I'd say the latter, if at all. I think updating access dates is an awfully long way down the list of priorities on which to spend one's time. It's dreadful how many editors seem to think access date is more important than, for instance, the actual publication date, in the case of a news reference. Access date is only ever an optional parameter, whereas the publication date of a news item is an essential piece of information. -- Alarics (talk) 11:29, 29 March 2011 (UTC)

I completey agree. There is nothing wrong with updating access dates, but it has a rather low priority.--Kmhkmh (talk) 11:40, 29 March 2011 (UTC)

The purpose of accessdate is give a starting point for recovery of the source if the link to an online source goes dead. The more up-to-date that date is, the more likely it is that Internet Archive will have a cached version of the page. That's why we don't bother with accessdate for online sources that are presumed stable. When you check an online reference, you could confirm that the page was available on that date. Whether it is a productive use of your time depends on how recent the current accessdate is and what priorities you assign to yourself. --RexxS (talk) 12:41, 29 March 2011 (UTC)

An additional use of the retrieval date is for sources that can be expected to change as new information becomes available. If you know that there have been recent changes in certain subject matter, and you notice the retrieval date precedes the changes, you should visit the source to see if it has new material. Jc3s5h (talk) 13:09, 29 March 2011 (UTC)

If you change the accessdate, you should make sure that the page at that date still has all the content(s) that it is used to cite. I recall one web reference that didn't contain anything related to the content it was attached in the wiki article at the time. I suppose some wiki editors would have just removed it, but the web page did have the content when it had been added to the wiki article initially, and an appropriately dated archive of the page had the content. Updating the accessdate for that web reference was not helpful. Gimmetoo (talk) 14:15, 29 March 2011 (UTC)

Thanks, that's another important point to consider. If the content of the online source has changed so as to no longer verify our article's text, then perhaps the url should be substituted with a link to the version of the page at the Internet archive that corresponds to the accessdate (and personally I'd leave the accessdate alone in that case). Obviously, it's important to check that the text in the article has not been superseded by different information in the meantime. --RexxS (talk) 17:10, 29 March 2011 (UTC)

You shouldn't really replace the content of the |url= parameter, instead you should add both |archiveurl= and |archivedate=, leaving |url= alone. --Redrose64 (talk) 12:33, 31 March 2011 (UTC)

In general I'd agree. But in the case of where "the content of the online source has changed so as to no longer verify our article's text", I'm not sure of the value of leaving in a link to that. YMMV. --RexxS (talk) 18:20, 31 March 2011 (UTC)

Thanks for your responses on this. I wasn't intending to go around randomly updating access dates for various articles, my query really was more related to if I was changing the citations anyway, and I take the point that if you do update the access date you should check that there is not more than one reference and that they are all still referenced at the source. Just to clarify one point above, access date is very important when there is no publication date listed for the source. I think that updating the access date may give some general confidence to the reader that the reference is still current, if it's been recently updated. Eldumpo (talk) 09:19, 30 March 2011 (UTC)

Some time ago there was an extended discussion of adopting a single citation style for Wikipedia. Does anyone remember where the discussion is located? Jc3s5h (talk) 15:32, 1 April 2011 (UTC)

Yeah. The same place that all failed proposals go. Please don't revive it: it provokes too much heated debate argument and never goes anywhere other than "no consensus - retain status quo". --Redrose64 (talk) 15:36, 1 April 2011 (UTC)

I was going to say whether or not I intend to revive it, but I won't, because the intentions of the requester is not an excuse for concealing information in a wiki. (Of course, Redrose64 might not know, off the cuff, where it is). Jc3s5h (talk) 15:45, 1 April 2011 (UTC)

I found it: Wikipedia:Centralized discussion/Citation discussion. Jc3s5h (talk) 16:14, 1 April 2011 (UTC)

That was 2010; the most recent that I'm aware of is Wikipedia talk:Citing sources/example_style#why_not_standardize_on_one_format.3F, from January 2011. WhatamIdoing (talk) 16:46, 1 April 2011 (UTC)

I was reading Win Butler's biography which states that he has met co-singer Regine Chassagne at McGill University. Other biographies on the band itself imply that they met at Concordia university.

Which is the correct version???

My Sources: Wikipedia —Preceding unsigned comment added by 178.152.52.22 (talk) 16:09, 3 April 2011 (UTC)

If I have several citations supporting a sentence, say "[1][2][3]", is there a way (a template?) to have this automatically be converted into "[1–3]"? bamse (talk) 10:12, 16 April 2011 (UTC)

I hope not! I think that would look terrible. -- Alarics (talk) 10:57, 16 April 2011 (UTC)

The Cite software does not support this. See WP:CITEBUNDLE, which simply moves clutter from the in-text cite to the reference list. ---— Gadget850 (Ed) ^talk 12:59, 16 April 2011 (UTC)

When you have refs of the form ^[1][2][3], each square-bracketed number is a separate link to a citation. If you try clicking such reference marks in Firefox, Chrome or Opera (but not Internet Explorer), you'll see that the target is highlighted in pale blue. Since the numbers need not be consecutive (see Me and Juliet (today's featured article), section Inception, paragraph 2 where you'll find ^[2][4] and also ^[4][6]), the links need to be distinct so that the correct target is reached. --Redrose64 (talk) 16:40, 16 April 2011 (UTC)

I understand that there are cases where reference numbers are not consecutive and also understand that each number is hyperlinked to the reference section. However there are many cases where reference numbers are consecutive and this method could be used in order to reduce clutter and improve readability. One could link "[1-3]" to all three references, i.e. highlight all three references when clicking on it. I am quite sure that I have seen such citation style in proper publications, so can't really understand the reservation to this proposal. I know there is a LaTeX style to achieve this (which does not help for wikipedia I know). bamse (talk) 17:14, 16 April 2011 (UTC)

See WP:CITEBUNDLE for a version of such a style, except it would just look like [1]. The problems with bundling is that it "simply moves clutter from the in-text cite to the reference list" and that if one reference in a bundle is used several times then serious clutter will ensue in the reference list. If you linked to the article you were working on, perhaps our outside eyes could help you with some style advice?  ⊂| Mr.choppers |⊃  (talk) 17:32, 16 April 2011 (UTC)

The reservation is that is technically not possible without a modification to the Cite software extension. Is there any style guide that uses this method? ---— Gadget850 (Ed) ^talk 18:09, 16 April 2011 (UTC)

re Mr.choppers: The article in question is this one. Cluttering is due to the way I create content, i.e. checking a number of references, creating notes of the most important facts and then compiling text out of it, transferring all the citations with it. So, often I end up with many references to the same sentence/fact. I guess, I need to get rid of the less important sources at some point. bamse (talk) 20:42, 16 April 2011 (UTC)

I think that it's seriously over-referenced. Do you really need eight refs for the sentence "It is a much more detailed version of the Kojiki, dating events and providing alternative versions of myths; it covers the time up to 697."? I also note that there are four red error messages in the references section. --Redrose64 (talk) 20:58, 16 April 2011 (UTC)

Fully agree with you. It is work in progress, so eventually (before nominating it at WP:FLC) I will get rid of some references. Will fix the red error messsages now. bamse (talk) 22:49, 16 April 2011 (UTC)

Wow! Yeah, going through and parsing the refs should take care of most of the issues you're having. By the time you're done with that, bundling the ones that are suited for it should leave you with a good looking article, without having to worry about contracting multople references in some new way. Bamse = Världens starkaste björn?  ⊂| Mr.choppers |⊃  (talk) 06:01, 17 April 2011 (UTC)

For an article I'm working on, there is a likely high-priority source that I need to use that is basically, at the present time, an e-featurette written specifically for the iPad (the work is discussed here: ). It is unlikely it will be made as a printed work, and also unlikely it would be posted to the web (the app has a website to tell people what it is about), but it may see other OS support (android, PC/Mac, etc), so it always will basically be an interactive application that provides information.

I cannot see any of the existing cite templates easily working for that. I can shoehorn it into them (cite web and pointing to the app's homepage would be one way), but I'd rather see if there's an existing template or should we consider one for the future if such types of works become more common. --MASEM (t) 13:12, 22 April 2011 (UTC)

I don't believe we have any cite templates, however to adequetly cite the source, I would suggest exploring traditional citation standards (like Turabian and MLA) and seeing if you can create a good and flexable citation with a lot of information, Sadads (talk) 13:21, 22 April 2011 (UTC)

Well, the article I've got (90% of it my contributions) uses templates like cite web, so for MOS consistency, I'd want to use what exists. The closest I can see from those without creating a new template would be cite web, with the "format" parameter set to indicate it as an e-magazine/feature. cite media would work as well, with the right "media" field. I can supply the usual things like author, publisher, date of release, etc. I'm more concerned if this style of journalism becomes common, if we need to consider a special template for that. --MASEM (t) 13:40, 22 April 2011 (UTC)

(edit conflict)Assuming this refers to Portal 2, then it already uses cite templates. Why not give us all of the pertinent info an let us see how it needs to be put together? ---— Gadget850 (Ed) ^talk 13:47, 22 April 2011 (UTC)

Although this may be quite obvious, we still need to have some formal policy on this (I am not sure if there already is. Please point me towards one if there is.). Very often different news reports from different reliable sources do not agree with each other. This happens more often when it comes to numbers (e.g. number of casualties, etc. See July_12,_2007_Baghdad_airstrike for example). Under such circumstances the best thing to do is to clearly mention that there are conflicting reports, and then clearly explain each report/view, and provide multiple citations for each. Please discuss this below. - Subh83 (talk | contribs) 18:32, 25 April 2011 (UTC)

Sometimes the estimates can change with time, so in those circumstances the best approach can be to use the latest estimates. Likewise, I've seen journal articles that provide each of the different estimates along with their authoritative sources. So I think it depends on the circumstances.—RJH (talk) 19:35, 25 April 2011 (UTC)

That's my understanding. As you said, mention both sides, explain what they think, provide very complete citations for each view.

This can be a difficult situation for an editor, because you have to balance two policies: WP:Neutral point of view and WP:Synthesis. If two sources disagree:

1. For WP:NPOV, you have to be careful to give both sides exactly the right amount of weight. If they are equally reliable, I think that WP:INTEXT attributions for both of them is appropriate. If one of the sources is central and the other is a fringe view, then you can use a footnote for central source, but you need an in-text attribution for the fringe view. Finally, you have to be very careful not to give too much weight to the side you personally agree with (or the side you wish was right, or the side you think is most interesting).
2. For WP:Synthesis, you have to be careful not to discuss the disagreement in any detail beyond what the sources actually say. Of course, sometimes it is difficult to find a source that says "sources disagree"; you often don't have a source for this statement. You have to try to present the disagreement without actually discussing the disagreement; for example something like "A said, B said" could be used for two equally reliable sources. But you have to be careful not to write too much about the disagreement itself. The more you analyze the difference, the closer you come to WP:SYNTHESIS.

This situation is one that requires intelligence, a solid understanding of Wikipedia's core policies and the integrity to implement them. ---- CharlesGillingham (talk) 22:53, 25 April 2011 (UTC)

Thanks for the pointer. I guess WP:Synthesis explains a lot what to do under such circumstances. Also, as RJH mentioned, it may be necessary to not give undue weight to outdated references. Indeed it seems to be requiring certain level of discretion. - Subh83 (talk | contribs) 20:39, 26 April 2011 (UTC)

My apologies if this has already been addressed and I just can't find it, but my question is what page number should be used when the source is a PDF file: The original page of the document or the page of the PDF file? The PDF file may also contain the document cover and other pages that are not numbered on a hard copy. Thanks, Alanraywiki (talk) 23:09, 7 May 2011 (UTC)

Use the page number the author inserted on the page— this will usually be the same as a hard copy version. ---— Gadget850 (Ed) ^talk 23:28, 7 May 2011 (UTC)

Thanks! Alanraywiki (talk) 23:30, 7 May 2011 (UTC)

The section Text-source integrity makes a claim that is not supportable:

The next section (Bundling citations) then goes on to suggest :

But to solve this alleged problem instead of bundling them one could simple add the additional information to the separate citations:

AFAICT there may be aesthetic reasons for bundling up citations but Text-source integrity is not one of the other reasons. If bundling is removed from Text-source integrity should the section be removed or should it be altered to say place inline citations where they most clearly support the text. and give an example of a longer sentence such as "Charels II's Scottish coronation took pace on 1 January 1651, while his English coronation was over 10 years later on 23 April 1661." with 2 citations at the end, and then, to demonstrate Text-source integrity the same two sources, one placed after the comma and the other at the end of the sentence. -- PBS (talk) 12:40, 17 April 2011 (UTC)

From my talk page:"Please stop reverting at policies and guidelines. SlimVirgin TALK|CONTRIBS 05:58, 21 April 2011 (UTC)"

From the history of the article: "05:57, 21 April 2011 SlimVirgin II (then pls discuss instead; some of the addition is unclear)"

I did not make the first first revert you did, and I have explained myself here (while you have been silent). How about explaining here what parts were unclear, and why revert all of the change if only some of it is unclear? -- PBS (talk) 06:13, 21 April 2011 (UTC)

Which claim is not supportable? SlimVirgin ^{TALK|CONTRIBS} 06:44, 21 April 2011 (UTC)

??? I do know that I posted a comment here three days ago and waited more than two days for a comment before making an edit to this guideline. I do know that that you made a revert of my edit without bothering to explain your reasons on this talk page -- despite this section existing -- and when I made a revert to your revert, instead of then posting an explanation here, with reasons for another revert, you post a request for me not to revert (something you do not seem to adhere to yourself), and instead of answering either of my questions you ask another question!

Coupled to my statement at the start of this section what parts of my edit to the guideline were unclear, and why revert all of the change if only some of it is unclear? -- PBS (talk) 07:02, 21 April 2011 (UTC)

SV you wrote "then pls discuss instead; some of the addition is unclear" If I am to discuss the changes with you need to explain what parts of my changes you think were unclear. -- PBS (talk) 15:55, 21 April 2011 (UTC)

Hi Philip, you said above: "The section Text-source integrity makes a claim that is not supportable ..." Can you say which claim is not supportable? SlimVirgin ^{TALK|CONTRIBS} 18:35, 21 April 2011 (UTC)

Philip, I also can't see which claim is not supportable. Jayjg ^(talk) 20:35, 21 April 2011 (UTC)

See above "But to solve this alleged problem instead of bundling them one could simple add the additional information to the separate citations: A simple example of not bundling citations:..." -- PBS (talk) 21:23, 22 April 2011 (UTC)

Philip, you wrote above, "The section Text-source integrity makes a claim that is not supportable." Please say what you mean (this is the third time I've asked). SlimVirgin ^{TALK|CONTRIBS} 17:34, 25 April 2011 (UTC)

I think he overstated it a bit, but it's clear what he means. The guideline currently suggests that bundling is the only way to provide text-source integrity in this situation. He's noticed that there is an alternative solution that doesn't involve bundling. So the guideline is wrong to suggest that bundling is the only solution. That's all. Isn't that clear? ---- CharlesGillingham (talk) 23:03, 25 April 2011 (UTC)

Thanks CG. SV when you last reverted my change you wrote in the edit history "The changes are not correct; pls explain on talk" yet all I have done is used an example from higher up the text (which is presumably "correct" (if not why is it in the guideline)) and then shown the same example with the citations at the end so breaking "Text-source integrity". What is incorrect with that? -- PBS (talk) 10:21, 26 April 2011 (UTC)

I'm sorry, I'm completely lost here. Philip wrote: "The section Text-source integrity makes a claim that is not supportable." Charles, the section doesn't say anything about bundling being the only way, so it's unlikely that's what he meant. Philip, can you say clearly which claim is not supportable, and quote the claim you're referring to?

Whatever he meant he has removed key material, and seems to have moved up material from the bundling section, with the result that the section is practically meaningless; hence my revert. SlimVirgin ^{TALK|CONTRIBS} 17:51, 26 April 2011 (UTC)

All the edit I made did to the Text-source integrity section was to delete "Where you are using multiple sources for one sentence, consider bundling citations at the end of the sentence or paragraph with an explanation in the footnote regarding which source supports which point; see below (Bundling citations) for how to do that." and used the previous example from the guideline to demonstrate a failure of in Text-source integrity

Bundling citations is not the only way that can done see the top of this talk page section for an example of another way it can be done. It can also be done by moving the citations back to where they are relevant as they are in the the original example sentences "The sun is pretty big,..." -- PBS (talk) 22:25, 26 April 2011 (UTC)

That wasn't all you did, Philip. You changed the section so it recommended sticking a footnote in the middle of a sentence. That's sometimes needed, but it's not something we should be recommending as standard, because it can often be avoided. And you removed the example. Also, can you please say what you meant when you wrote above: "The section Text-source integrity makes a claim that is not supportable ..."? SlimVirgin ^{TALK|CONTRIBS} 22:32, 26 April 2011 (UTC)

the false claim is "see below (Bundling citations) for how to do that" as it can be done in other ways and that is not the only way. -- PBS (talk) 23:01, 26 April 2011 (UTC)

The example I used is one from higher up the page, so the example I used is internally consistent. If it is wrong in this section then it is wrong higher up the page. Besides what is wrong with placing a citation after a clause in a sentence? -- PBS (talk) 22:46, 26 April 2011 (UTC)

It says "consider doing X," see below. That doesn't say or imply that it's the only way, and that's anyway not what's meant by an unsupportable claim. Philip, this isn't helpful.

As for citations inside sentence, people generally try to avoid them, but as the guideline says, they're allowed. SlimVirgin ^{TALK|CONTRIBS} 22:13, 27 April 2011 (UTC)

I think that the consider sentence coupled with the wording in next section is a false claim. But I an not going to argue further with you over whether it is claim or something else. It is that the wording is inappropriate for this section, as it is only one of several ways to solve the problem, and there is no need in this section to emphasise one solution. It can be deleted without affecting the important message conveyed in the section. As to your point "people generally try to avoid [placing citations after commas]" I suspect you mean "I (SV) generally try to avoid them". Perhaps for the sake of "text-source integrity" you should reconsider that self imposed prohibition.

As the Sun example is used in previous sections it seems to me appropriate to use it here to demonstrate text-source integrity by presenting it as it has already been presented and then again with the citations all clumped together at the end of the two sentences. -- PBS (talk) 16:38, 28 April 2011 (UTC)

There you go again, Philip. I didn't say people should avoid placing citations after commas, but that people generally try to avoid placing them inside sentences, because^[5] it^[6] looks^[7] bad,^[8] is rarely if ever found outside Wikipedia, and is never necessary. SlimVirgin ^{TALK|CONTRIBS} 17:08, 28 April 2011 (UTC)

Exaggerating to make a point? I don't think it makes the point well. People can read what we both wrote above and decide who said what, and come to their own conclusion, so lets drop the argument, and concentrate on the changes I wish to make and you reverted.

The example I quoted in my edit to the section was not one I made up, but one I copied from another section higher up the page called Footnotes, so if you think it inappropriate here then presumably you think it inappropriate in the footnote section as well. If so, then get agreement to change it in the section footnotes (Personally I think it is a perfectly good example), as I am happy to use whatever is the example given in the section footnotes in this section and then show the same example with all the footnotes lumped at the end of the example to demonstrate a break in Text-source integrity. -- PBS (talk) 16:47, 29 April 2011 (UTC)

From the history of the article:

• 15:33, 6 May 2011 Philip Baird Shearer (Explanation given on the talk page so putting back new example for Text-source integrity and bundling)
• 01:26, 7 May 2011 SlimVirgin (was better before; not clear what some of this means)

"was better before;" matter of opinion. "not clear what some of this means" Which part? -- PBS (talk) 09:29, 7 May 2011 (UTC)

(Query about a specific problem moved to user's talk page and replied to there.  —SMALLJIM  12:57, 10 May 2011 (UTC))

I have been working on User:Gadget850/Citation templates— anchors. Should this be moved into Wikipedia space somewhere so it can be used and updated? ---— Gadget850 (Ed) ^talk 14:27, 5 May 2011 (UTC)

I support moving it to the mainspace. Try asking at WP:VPP for where, exactly. --_{Piotr Konieczny aka Prokonsul Piotrus| talk} 18:26, 13 May 2011 (UTC)

I have seen short footnotes used effectively where a book is cited at several different pages. Each page ref has a separate footnote, and then the full citation is given in the "References" or "Bibliography" section. However, I am starting to see a number of short articles now that do not have multiple footnotes to different pages in the same book, and can be easily covered by the full citation in the footnotes without a separate references section. I find this to be much easier to use because the footnotes have a ^ link symbol that allows me to return to the text, which is not present in the bibliography. Could we amend the short footnote MOS to state a preference for the long footnotes except in cases where multiple footnotes refer to different pages in the same book or other source? Thanks Racepacket (talk) 18:02, 7 May 2011 (UTC)

No: The anti-holy war rule is that editors may use any system they choose, by which we mean any system, including systems that I personally don't like. If the editors at an article choose to make zero citations clickable, or even to manually number the citations (*shudder*), then we leave them free to do that.

But you might like to read at bit more about some of the creative solutions people have used for formatting citations, because there's nothing in the system you describe that prevents a fully clickable system. WhatamIdoing (talk) 18:50, 7 May 2011 (UTC)

I am all for maximum freedom, including parenthetical references in the text rather than in footnotes. However, if footnotes are used, can we maximize the reader's ability to return from the source back to where he/she was reading in the main text. The ^ link in the CITE extension is invaluable, and something similar is needed in the alternative referencing schemes. Racepacket (talk) 19:51, 15 May 2011 (UTC)

If you linked from the in-text cite to the shortened footnote, then the long footnote, the pressing backspace twice should return to the in-text cite. ---— Gadget850 (Ed) ^talk 23:10, 15 May 2011 (UTC)

An issue came up during the GA review for Don't Look Now in regards to the reference format. The review is at Talk:Don't_Look_Now/GA1, and the relevant parts are the reviewer's last review point and my point listed at #8.

The issue revolved around me putting the website name in the 'work' parameter:

Canby, Vincent (10 December 1973). "Don't Look Now (1973) – Film:'Don't Look Now,' a Horror Tale:Donald Sutherland and Julie Christie in Leads The Cast Suspense Yarn Turns Into a Travelogue". nytimes.com. The New York Times. Retrieved 21 February 2011.

The reviewer felt that the newspaper name (The New York Times) should have gone in the 'work' parameter and the "The New York Times Company" in the 'publisher' parameter.

I'm happy to do this if this is the preferred protocol, but the reason I did it is that newspaper references rarely list the publishing company, they effectively publish themselves. In all the academic references I've seen only the newspaper name is given. When it's sourced online it is usually accompanied by the full url, so the name of the website is still explicitly given in the reference, which is effectively where you got your information from. A quirk of Wikipedia is that the url is hidden beneath the title, so just to make the information sources explicit I used the work parameter to provide the website address; it just seemed a bit weird looking at list of online references and not seeing the website names.

There are a lot of policies and guidelines about referencing so my apologies if this is covered somewhere, but is there a specific guideline for using these parameters in this context? I can pull the work parameter if I've incorrectly used it, that isn't a problem, but is it common practice to add in the actual publishing company? My aim is to get the article to FA status eventually, so I'd like to get the references in order, especially if my citations are unorthodox and could derail an FA nomination. I'm interested in personal opinions as well as the policy perspective. Betty Logan (talk) 05:31, 2 April 2011 (UTC)

There is no need to put nytimes.com anywhere. "The New York Times" goes under "work" (or, better perhaps, "newspaper") -- either of these will ensure that the name of the paper appears in italics, as required by convention -- and the "publisher" parameter can be deleted altogether. "Publisher" is only ever optional, and is not needed for a mainstream newspaper -- it is meant to be used when an obscure local newspaper is published by a wider newspaper group. There is no need to say who publishes The New York Times. See Template:Cite news. -- Alarics (talk) 07:31, 2 April 2011 (UTC)

Thanks for replying. I will pull the obsolete information in that case. Betty Logan (talk) 01:43, 6 April 2011 (UTC)

I apologize for the late reply, but:

The publisher of The New York Times is Arthur Ochs Sulzberger, Jr.. The owner is The New York Times Company. In the newspaper industry, publisher is a job, with specific tasks to be done by a human (like the executive producer of a movie). WhatamIdoing (talk) 18:20, 7 May 2011 (UTC)

True in America, not generally true in the UK, where the appearance of "publisher" in that sense is very recent. Traditionally, and mostly still, "publisher" in the UK means the publishing company, and the purpose of this parameter in "cite news", as I understand it, is for where an obscure local newspaper (which does not have its own Wikipedia article, so cannot be wikilinked) is one of a chain of papers published by a larger group, in which case one can put the name of the larger group under "publisher". It is usually expressed somewhere on the title page of the publication concerned as "published by X Ltd of X". I cannot see that one would ever need to cite the publisher in the American sense of the word. (For the avoidance of confusion, I would be very happy to see this near-useless parameter abolished.) -- Alarics (talk) 19:39, 7 May 2011 (UTC)

That is not exactly true. In the UK, the job title of "publisher" for the non-editorial business manager is long and well established in the magazine & small paper sector, but causes confusion outside the industry. In the template context, "publisher" clearly relates to books, & should be left blank for newspapers, imo, at least large ones, though the owning company could be added for magazines & small papers that are in a large group. Johnbod (talk) 19:38, 13 May 2011 (UTC)

Or perhaps it should be replaced by "owner". I can imagine situations in which naming the conglomerate that owns the paper would be appropriate, e.g., a Hearst publication writing about Mr Hearst. WhatamIdoing (talk) 20:49, 7 May 2011 (UTC)

It would be good if a note could be left on these templates telling people that the publisher parameter isn't needed for newspapers. First, if you put the name of the paper in that parameter it's not italicized as it would be as "work." More importantly it introduces extra clutter and makes people think they have to supply it. I've seen several newbies mess around trying to find the name of the publisher, thinking it was required. SlimVirgin ^{TALK|CONTRIBS} 19:46, 13 May 2011 (UTC)

Indeed so, and I (being one of the relatively few people on WP who pay attention to getting references right) must have spent, or wasted, hundreds of hours putting such errors right. In the case of "cite news", can we not simply abolish the "publisher" parameter? -- Alarics (talk) 22:47, 13 May 2011 (UTC)

I've left a note on the template talk page asking if I can remove it. SlimVirgin ^{TALK|CONTRIBS}

And if we can't remove it (or rename it to "owner" or something more obvious), then perhaps we should ask for a bot to correct the use in the most commonly cited media outlets. WhatamIdoing (talk) 04:37, 17 May 2011 (UTC)

I am doing some cleanup on Fantastic Mr. Fox (film) and encountered the use of a wikiquote as a references. The existence of {{Cite wikisource}} suggests it is generally accepted to cite wikisource; is it also okay to cite wikiquote, and if so, is there a template that is recommended to use? Thanks. 67.100.125.30 (talk) 01:36, 17 May 2011 (UTC)

{{Cite wikisource}} isn't really a citation template— it is more of an external link template. Wikiquote only supports that the statement was made in the movie, the second source from TV Tropes makes the comparison. ---— Gadget850 (Ed) ^talk 03:11, 17 May 2011 (UTC)

I think that for high quality articles, FA at the very least, all book references should have Google Book page links (where possible). They make verification tremendously easier. Thoughts? --_{Piotr Konieczny aka Prokonsul Piotrus| talk} 18:24, 13 May 2011 (UTC)

We had an RfC on this not long ago, and it was decided they could always be added (so long as they were page links, not general links), and that editors should not remove them, but they shouldn't be mandatory. SlimVirgin ^{TALK|CONTRIBS} 18:41, 13 May 2011 (UTC)

If you do include a Google book lookup url, as a best practice I'd like to suggest that it be limited to the 'id' and 'pg' name/value encoded pairs. Thus:

url=http://books.google.com/books?id=beRzpeiXod0C&pg=PA50

rather than:

url=http://books.google.com/books?id=beRzpeiXod0C&pg=PA50&dq=%22outer+space%22+definition&hl=en&ei=UYDNTcSQE4j0tgOQ6dDDCw&sa=X&oi=book_result&ct=result&resnum=1&ved=0CEYQ6AEwAA#v=onepage&q=%22outer%20space%22%20definition&f=false

Everything else just seems to add a lot of bulk to the article download. Thanks. Regards, RJH (talk) 19:10, 13 May 2011 (UTC)

Thanks, agreed. There's a section in CITE explaining how to write the URLs. SlimVirgin ^{TALK|CONTRIBS} 19:33, 13 May 2011 (UTC)

I am always a bit reluctant to add GBOOKS urls to cited books in an article. Doesn't linking to such a controversial web service violate WP:NPOV? --Eisfbnore ^talk 20:33, 13 May 2011 (UTC)

I don't see how it violates NPOV any more than it would citing a hardcopy edition of the book. Betty Logan (talk) 20:51, 13 May 2011 (UTC)

It's quite easy: just as we don't put links to Google, Yahoo! or Bing Maps in EL sections – but rather let the readers decide themselves what map service they'd like to use through the geo-coordinates – should we not include links to Google or any other non-Wikipedia site in the book sources. Better to simply leave the book title plain and let the readers choose for themselves how to locate it with the help of the ISBN or OCLC. Eisfbnore ^talk 10:38, 18 May 2011 (UTC)

But people do add links to Google Books to EL sections; there's no reason not to. This approach of removing helpful material just because it's commercial is very wrong-headed. Take that approach to our source material and watch the project grind to a halt. SlimVirgin ^{TALK|CONTRIBS} 11:53, 18 May 2011 (UTC)

It takes exactly three clicks from the article to the GBOOK (ISBN → Special:Booksources → "Find this book at Google Book Search online database" → voila). Additionally, they're not always accessible for everybody around the globe, just see Wikipedia:Featured article candidates/Black Friday (1945)/archive1 for a discussion of this. --Eisfbnore ^talk 12:10, 18 May 2011 (UTC)

Why force people to click three times when they need only click once? Google Books pages are widely available. We don't argue for the deletion of material not available in China, so if editors in some countries can't see them, that's no reason to affect editors who can. Anyway, we had an RfC about this. No need to rehash it here. SlimVirgin ^{TALK|CONTRIBS} 12:15, 18 May 2011 (UTC)

There isn't really much to debate here. Google links are optional but of course not mandatory. There is no requirement for sources being available online nor is there (currently) a consent for such a thing iamginable.--Kmhkmh (talk) 21:15, 13 May 2011 (UTC) Also note that if you want to use Google links nevertheless, that there is a template available: {{Google books|ID|displayed text|page=}}--Kmhkmh (talk) 21:17, 13 May 2011 (UTC)

Applies to other hosts such as Scribd. Agree with the trimmed links. If you link to a page, then you can reuse the reference only if it is to the same page, references to other pages would have to be separate. Links to bokks with no view or snippet view is not useful. Snippet view is very problematic as it looses context. ---— Gadget850 (Ed) ^talk 22:25, 13 May 2011 (UTC)

Actually, I've not heard of Scribd having an agreement with anyone, so I remove those links as most likely being a contributory copyvio. Tijfo098 (talk) 20:23, 19 May 2011 (UTC)

Also, one should point out that the legal controversy is about the GB books with "snippet view", which are fairly irrelevant here because you can't really link to individual page of those. Those with (limited) page view have an explicit agreement with the publisher. See WSJ; I've updated the Wikipedia article from that. Tijfo098 (talk) 20:19, 19 May 2011 (UTC)

http://faq.dtnorway.com/question?questionid=1032#comments —Preceding unsigned comment added by Atremist (talk • contribs) 09:57, 21 May 2011 (UTC)

You should raise individual article related isuues on it's talk page. —  HELLKNOWZ  ▎TALK 11:11, 21 May 2011 (UTC)

A relevant RfC is in progress at Wikipedia:Requests for comment/Dead url parameter for citations. Your comments are welcome, thanks! —  HELLKNOWZ  ▎TALK 10:59, 21 May 2011 (UTC)

Hi all: I thought it might be useful to have a note in "Identifying parts of a source" about using times to refer to specific points of interest in audio-visual material. I moved the "books" example into its own bullet point and expanded it to "books and print articles", and added a new bullet point for "audio and video sources". The various {{cite media}} templates support this parameter already.

I also added "Sound recordings" and "Film, TV, or video recordings" to the "Examples" section to indicate the kind of information that usually goes into citing this type of material.

More eyes and fine-tuning certainly welcome! Cheers --Rlandmann (talk) 12:16, 21 May 2011 (UTC)

This is a common misconception. I've seen this reasonably enough to wonder if we shouldn't make some clear statement in this (?) policy that yes, most sentences need references. With a nod towards Wikipedia:You don't need to cite that the sky is blue, this is a wiki, and sentences may be moved, added to, and so on. Just few days ago I reviewed a DYK and asked the nom for an ref for the hook inside a para. The para's last sentence had a same ref, and you'd think the middle, where the hook was from, would have the same one, right? But it was instead another source the nom used, but forgot to add to that para (see this). This is just one of many cases which prove that it is better safe than sorry, and referencing each sentence is the right thing to do.

If you are not convinced, imagine a paragraph with only end-of-para reference. Do you assume that the last reference cover all the sentences? If so, you trust Wikipedia more than I would. Can you be certain that it wasn't an unreferenced para to which somebody added a new, referenced sentence, or if that ref at some point really was for all sentences, that nobody added unreferenced content to the middle? If somebody were to add new content to the middle of that para, with a reference - would you then assume that the beginning of that para is referenced with the new, middle-of-the para ref? Or if such a para was split into two, thus leaving the first one unreferenced? I hope it is clear why each sentence has to have its own reference. Suggestions on how to stress it (in this policy?) would be welcome.

PS. A while ago, I reviewed a DYK and asked the nom for inline cites. He grudgingly added the single one necessary for the hook, and for the few cite needed templates I added. The article used a single source, and he "promised" me that he would remove all the inline cites as "the article is good enough with a general reference, and all those footnotes look ugly...". Ugh. --_{Piotr Konieczny aka Prokonsul Piotrus| talk} 01:00, 1 June 2011 (UTC)

You raised this here not long ago, Piotrus, and the situation hasn't changed. Most sentences don't necessarily need references after them. It depends on the context. Personally, I like to see at least one reference after each paragraph, with the citations bundled, so readers don't have to hunt too far for the sources. But there are editors who argue that a refererence after each paragraph isn't always necessary either. SlimVirgin ^{TALK|CONTRIBS} 01:04, 1 June 2011 (UTC)

The last time I checked, WP:V was a policy. Did it get demoted? PS. It seems that at each stage we try to improve verifiability of our articles, there is a vocal minority who scream "murder". How long did it take us to make inline referencing a standard? At one point, the single "general reference" was more than enough. Just like this is now a piece of wiki history, so does the "one ref is good enough for a para" fallacy must be put in the history bin, for the good of this project. --_{Piotr Konieczny aka Prokonsul Piotrus| talk} 01:09, 1 June 2011 (UTC)

V doesn't require a citation after every sentence. Nor does CITE. Regarding this edit of yours, it's you that's in the wrong there, I'm afraid. There's no need to keep repeating the same reference after each sentence like that. SlimVirgin ^{TALK|CONTRIBS} 01:13, 1 June 2011 (UTC)

Bang head on the wall. And how soon till somebody splits the paras, or adds unref information? Case to point, are you saying I was in the wrong when I asked for a ref that resulted in this?

But you are correct we may never reach a consensus just talking, last time we talked for quite a while and with the usual little effect. I think it is time for a poll, to close this issue for at least few more months and with less ambiguity. --_{Piotr Konieczny aka Prokonsul Piotrus| talk} 01:16, 1 June 2011 (UTC)

(ec) Look at your edit in read mode. The repetition of the same footnote hasn't added anything that would help a reader:

"By mid-1944, the humanistic department had 32 lectures and 168 students..^[9] The medical department, created in the 1942/1943 year, was one of the largest, with 610 students and 95 lecturers in the 1943/1944 year..^[9] The theology department had only 5 lecturers and 20 students..^[9] The law department was the most popular; the year 1943/1944 had over 600 listeners..^[9] Overall, in the 1943/1944 year, the university had about 1,200 students..^[9] Throughout its period of operations it awarded 95 master degrees, 5 doctoral and 5 habilitations..^[9] Plans to create new departments were obsoleted by the Uprising.^[9] An underground Lawyer Club was also created, holding active presentations and discussions..^[9] This department awarded 54 master and one doctoral degree.^[9] "

SlimVirgin ^{TALK|CONTRIBS} 01:18, 1 June 2011 (UTC)

So you think this would be more helpful?

"By mid-1944, the humanistic department had 32 lectures and 168 students.^{[citation needed]} The medical department, created in the 1942/1943 year, was one of the largest, with 610 students and 95 lecturers in the 1943/1944 year.^{[citation needed]} The theology department had only 5 lecturers and 20 students.^{[citation needed]}The law department was the most popular; the year 1943/1944 had over 600 listeners.^{[citation needed]} Overall, in the 1943/1944 year, the university had about 1,200 students.^{[citation needed]} Throughout its period of operations it awarded 95 master degrees, 5 doctoral and 5 habilitations.^{[citation needed]} Plans to create new departments were obsoleted by the Uprising.^{[citation needed]} An underground Lawyer Club was also created, holding active presentations and discussions.^{[citation needed]} This department awarded 54 master and one doctoral degree.^[9]"

--_{Piotr KonieczngetmasksimBitmap_Commandy aka Prokonsul Piotrus| talk} 01:20, 1 June 2011 (UTC)

Piotrus has just arrived at an article I've been editing, and that he has never edited, to add an unreferenced tag because of this exchange. This is POINTy and disruptive. Please don't do that kind of thing again. SlimVirgin ^{TALK|CONTRIBS} 01:23, 1 June 2011 (UTC)

Oh please. Slim has arrived at an article I've been editing, cited it here and that's fine (and it is, I am not complaining), but if I dare to look at her contribs and suggest one policy-driven improvement tentatively related to this discussion that's WP:POINT? And so you revert my constructive edit because of this discussion? Should we now start edit warring there like children? How about you restore my edit there (or add requested citation), and answer my questions above that you've been ignoring (re: my post from 01:16, 1 June 2011) instead of changing the subjects into personal attacks? --_{Piotr Konieczny aka Prokonsul Piotrus| talk} 01:28, 1 June 2011 (UTC)

You linked to your article at the very top of this section. I wouldn't otherwise have looked at it. This is my last post here, because it's clear this is going nowhere. SlimVirgin ^{TALK|CONTRIBS} 01:33, 1 June 2011 (UTC)

So you refuse to answer my simple question above ("are you saying I was in the wrong when I asked for a ref that resulted in this"), or the one from 01:20, 1 June 2011, and instead end up your contribution by reverting my WP:V-supported citation request to an article I tried, in good faith, to help you out and improve, accusing my of POINTing. I hope you'll feel more collaborative in a few days, no hard feelings. I will also have no problem if you'd like to refactor this discussion, collapsing this entire discussion at this point. PS. For future reference: anybody is always welcome to edit articles I was/am editing, leave tags, comments and any other requests for further improvements. Whenever my comments on some talk page inspire anybody to review my comments and join in collaborative editing, I consider it a very good day :) --_{Piotr Konieczny aka Prokonsul Piotrus| talk} 01:39, 1 June 2011 (UTC)

Sourcing single sentences rather than paragraphs might have a slightly better text source integrity but you can't trust that either and it can easily manipulated by other editors as well. Hence the notion that single sentence sourcing is always a real improvement is imho nonsense and false focus on formalities or visual gimmicks. The only way to check whether an article is properly sourced is to actually read the sources. Generally we require only that the citation is "reasonably close" to the sourced text, which you usually means that latest at the end of the paragraph the source should show up or in the case of very short articles you simply might list them at the end. If you want to be particularly precise yourself you could use annotated bundled citations. However I don't think that is something can be required (maybe recommended though), as it significantly increases the workload for the author and as a general idea keep it simple to contribute.--Kmhkmh (talk) 01:25, 1 June 2011 (UTC)

I tentatively agree that it should be suggested at most levels, and required at higher end. Whether to require it only at FA or below I'll leave to others to hash out; personally it is something I expect from GA-class at the very least. I am fine with adding it as a suggestion to this policy, rather than a requirement. What I am not so fine with is people saying that it is quite unnecessary at all, and "one ref per para" is "generally enough". --_{Piotr Konieczny aka Prokonsul Piotrus| talk} 01:31, 1 June 2011 (UTC)

Honestly I really don't care much whether it is required for FA or not (though personally I consider strict single sentence sourcing as ugly and somewhat pointless), but this here is guideline for general articles and therefore not really the place for FA criteria, there's a separate project page for that.

Imho it is rather unproductive if we have people running around and doing with paragraph based citations, what you just did with the example above. In my eyes that serves no purpose at all and the only valid reason to add a "citation needed"-tag to a sentence there is when you have good reasons to believe that one of the sentences is not covered by the source. What we don't need really need people checking whether every sentence has an inline citation, what we do need is people actually reading the sources to check whether the content is correct or not.--Kmhkmh (talk) 01:44, 1 June 2011 (UTC)

Sadly, we will not get this till there is some easy mechanism to indicate whether people have reviewed a reference. I'd love to see some "verified" mechanism for templates, maybe in the form of Facebook likes, but it is not something we have, and not something we are likely to have in the near future. In the absense of that, we have to focus on improving the articles in the current system, and I think that can be easily achieved by putting more stress on referencing all "non-blue sky" sentences. With my edit above, you no longer have to worry whether those sentences are or aren't referenced by the end-of-para ref; you can be assured they are (or at least, that somebody believes they are). This makes those sentences more trustworthy than if they didn't have those cites, and till we get the "verified by many editors" tool for reference, it is as good as we can get. --_{Piotr Konieczny aka Prokonsul Piotrus| talk} 01:52, 1 June 2011 (UTC)

(ec) I agree with Kmhkmh, and disagree very sharply with Piotrus. Wikipedia has moved much too far in the direction of citing every sentence. It's ugly, formalistic, and tends to disrupt the narrative. --Trovatore (talk) 01:55, 1 June 2011 (UTC)

The answer's still no. It is the job of any editor who splits a paragraph to make sure that any relevant citations are handled correctly. Normally, this means a quick copy-and-paste job, but it is the subsequent editor's job to do this, not the original editor's.

Wikipedia does not mandate citations on the basis of grammatical constructs. Some sections require zero inline citations. Frequently, one citation is enough for a multi-sentence, single-point paragraph. Some material requires multiple citations per sentence. "One citation per sentence" is not, and should never be, the 'rule'. WhatamIdoing (talk) 06:04, 1 June 2011 (UTC)

The important thing to remember here is that citations are primarily for editors and "power readers", the vast majority of people aren't interested in finding out where the material came from.

Our policies are clear; every piece of material must be verifiable, but there is nothing that prescribes whether that is via inline citation, bibliography or simply by linking to it in an edit summary. This is good, it gives editors a degree of flexibility, which means that if you are writing an article and want to cite every single sentence then you can do so. And those of us who prefer to use citations in a more restrained way can do so.

There are some Wikipedia processes which force a particular form of reference; for example if you want an article to be promoted for FA (and to a lesser extend GA) you are going to need to have inline cites at least for each paragraph (and, conversely, you'd likely be asked to remove cites duplicated sentence after sentence).

On the specific issue of citing each sentence; I do think it is messy, and I suggest that we try to be considerate of readers when placing cites. It is the same issue of "citation stacking" where a word or sentence is queued up with 5 references to support it (perhaps because it has been contentious). This can interrupt the flow of reading.

When I read an article (for my own education/enjoyment) I usually hide the references (for which I have a neat little script) because it can become distracting.

Reading Piotrus' comments I have to question what purpose citing every sentence serves; everything has to be verifiable, whether it has an inline citation or not. I do not think that having a duplicated inline citation for every sentence helps - if those citations are all different, with specific page numbers, then yes, perhaps. But otherwise surely a paragrapgh sourced to a single source need only be cited once :)

There is a balance between editorial ease and readability, I think demanding per-sentence citation for every sentence sacrifices too much readability. For exmaple, my pet article would be a mess of ^[1]'s and each reference at the foot of the page would have about 20 links^{abcdefghijklmnopqrstuv}. I fail to see the use of this.

Essentially this is making the argument that stuffing inline citations improves the referencing of articles, a concept which completely misses the point IMO and does nothing to improve quality, verifiability or readability. Everything should be verifiable to a source, ideally that source should be noted on the article and convention suggests adequate inline citations for ease of verification. Requiring per sentence citations is a step way too far (it is something you would never get past the FA crowd, for example) and serves no useful purpose.

On a personal note; if I came across a paragraph with every sentence cited to the same sources it would actually raise a red flag to me as likely dubious content that needs to be looked at. --Errant ^(chat!) 10:27, 1 June 2011 (UTC)

Personal opinions - If an article has sentence by sentence references, it is very difficult to read. Not strictly "unreadable" but close enough. And if editors were actually required to provide the amount of referencing advocated by Piotrus, many would give up editing. Wanderer57 (talk) 14:41, 1 June 2011 (UTC)

Yes, I agree. I've never seen a scholarly paper that cites every sentence. In Wikipedia, once per paragraph is usually more than enough to confirm that an article has been at least decently researched. Scholars of the topic can then drill down and confirm (or tag) specific statements. Regards, RJH (talk) 14:46, 1 June 2011 (UTC)

I'm with the majority here - especially as I am currently in a year-long battle with an editor who steadfastly produces sentences such as these:

With production bases located in China,[9] the Philippines,[9] and Taiwan,[9] Yulon makes license-built[citation needed] versions of many automakers' models.[9] The companies it manufactures in cooperation with include Chrysler,[9] Geely,[10] GM,[9] Mercedes Benz,[9] Mitsubishi,[9] and Nissan.[9]

(from an old revision of Yulon Motor)  ⊂| Mr.choppers |⊃  (talk) 14:54, 1 June 2011 (UTC)

A bit of an extreme example, but I can see both sides of the coin. If a sentence is cobbled from multiple refs, this is understandable, although bundling citations could be a solution. The advantage of the style you cite is that it makes it most easy to verify the particular claim; the disadvantage, of course, is clutter affecting readability. Maybe a solution would be, for those annoyed by the clutter, to be able to "turn off" footnotes visibility? --_{Piotr Konieczny aka Prokonsul Piotrus| talk} 17:34, 1 June 2011 (UTC)

Rather than having the reader guess what part of the text is cited, it would be nice if the range were highlighted in some manner. A possibility would be to change the shape of the cursor over a range of cited text. Currently the page shows either the 'text' cursor or the 'hand' cursor. What if it showed, say, the 'default' (or perhaps the 'help') cursor over a range of cited text? This would provide feedback to the viewer. For example, this range of blue text has a different cursor shape.^[1] Regards, RJH (talk) 16:23, 1 June 2011 (UTC)

In an ideal world we could have the current system of citation - but with a editor-focused system of referencing and highlighting that lets you make more specific markers (that don't appear in the normal page display) to help with verification, referencing and notation. Something that could be turned on to note "this conflates material in chapter 2 & 3" etc. --Errant ^(chat!) 16:35, 1 June 2011 (UTC)

I, for one, am not sure which universe the folks who oppose referencing each sentence have arrived from—have they not edited anything around here or what? It would, of course, be great if all editors always made sure that whatever they are editing at the moment is covered by the source which is cited at the end of the paragraph, but let's be real, shall we? It's not gonna happen. Ours, indeed, is not an ideal world. People move stuff around, insert/delete sentences, and re-arrange citations all the time, and more often than not all they care about is the material, not how well that material is represented in the source being cited. Sadly, for many people referencing is an afterthought, a mundane chore, something to get out of the way quickly. And while such attitudes continue, it's very naive to expect the ideal behavior from everyone. Ugly or not, citing every sentence is the best way to keep up with the fluidity of the content, and while not everyone wants to do it or is even capable of doing it, at least don't tell those who are to cease and desist!—Ëzhiki (Igels Hérissonovich Ïzhakoff-Amursky) • (yo?); June 1, 2011; 17:20 (UTC)

I think all/most editors here are aware of the editorial "misbehaviour" as you describe it. The question however is how and whether than can be fixed and I don't quite see how single sentence sourcing is providing any substantial improvement here. Moreover I'm getting more and more the impression that single setence sourcing is favouring another editorial "misbehaviour" or misconception on its own, in particular that people associate single sentence sourcing with correct content or "reliable sourcing". You cannot judge whether something is properly sourced by checking whether it has single sentence sourcing but you actually need to read the sources. And if you do read the sources, it makes hardly any difference whether they are sentence based paragraph, based or more generally just close enough. Even if you have single sentence sources people can (and will) simply borrow the sources from the already existing sentences for their new lines. They can introduce new stuff by modifying/rewriting sentences, adding half sentences and leaving the sourcing untouched. In other words single sentence sourcing is primarily a cosmetic operation which in addition signals a false sense of security.--Kmhkmh (talk) 20:41, 1 June 2011 (UTC)

I, for one, have no thought of telling someone of the "sentence by sentence citation" persuasion to cease and desist. My comment above was made out of concern that some editors were so convinced of the importance of the "sentence by sentence citation" approach that they want to tell other editors that it MUST be followed.

Editor Ëzhiki: I found this sentence of yours ("I, for one, am not sure which universe the folks who oppose referencing each sentence have arrived from—have they not edited anything around here or what?") insulting & uncalled for. I would think someone with as much editing experience as you seem to have would know that Wikipedia editors vary hugely in their background, experience, interests, methods, etc. and that expecting them to conform to your practices is futile at best. Wanderer57 (talk) 18:10, 1 June 2011 (UTC)

Citizen Wanderer, if that sentence of mine found you offended, I, sadly, cannot help you. One picks the sentences to be offended with on one's own; I have no control over that process. For what it's worth, the sentence was intended more as a lighthearted introduction (one which is based on numerous observations of real-life editing, by the way, hence the tinge of surprise), but if you personally found it offensive, please accept my apologies. That taken care of, as far the rest of my comment goes, how in the world did it give you an impression that I "expect... [editors] to conform to [my] practices"??? All I expect other editors to do is not to mangle the sentence-by-sentence references when they are already done (unless there is a discrepancy between what the article says and what the sources say).—Ëzhiki (Igels Hérissonovich Ïzhakoff-Amursky) • (yo?); June 2, 2011; 15:30 (UTC)

You know, I asked for general opinions a while ago (a year ago?) at one of the FA talk pages, since the folks there follow the highest standards. I had expected something like "well, we never pass an article that doesn't have at least one inline citation per paragraph", so I was just a bit surprised to see that the consensus was to include every inline citation that was necessary to show where the material came from, but no more than that.

Upon reflection, I decided that it was actually a good standard. IMO it is a much better standard than "one per grammatical unit", especially if you assume that our editors are capable of using good judgment. Mindlessness is not a desirable trait for editors (or policy writers). WhatamIdoing (talk) 19:48, 1 June 2011 (UTC)

I agree with the consensus here and policy, that one need not (and indeed should not) cite every sentence unless required. A citation that covers two or three sentences, or even an entire paragraph, is preferable to filling up the page with repeated citation superscripts. Jayjg ^(talk) 23:25, 1 June 2011 (UTC)

Ëzhiki, I think it is important have experience with a problem before proposing a solution. Citations getting mixed up is a problem - it is not one caused by citing paragraphs, and it is not one solved by citing sentences. I've worked on articles where citations get churned all over the place regardless of where they were placed. If the article is undergoing a lot of work by a number of editors you are going to lose track of sources, nothing much can change that. If you have an article with minimal activity where an editor comes in an breaks up a paragraph then it is their responsibility to make sure it is all adequately cited afterwards :) Sure, citing sentences allows them to be lazy, but it's too much of a trade off to disrupt the reader. --Errant ^(chat!) 10:11, 2 June 2011 (UTC)

I realize that full well. However, from my experience, when an article is undergoing a lot of work by a number of editors, it is much more difficult to lose track of the sources when the individual sentences are referenced as opposed to when a reference is supposed to cover a whole paragraph. If there were an easy way to determine which parts the supplied reference covers, then sentence-by-sentence referencing would indeed be redundant. In the absence of such an easy way, referencing each sentence is the safest bet.—Ëzhiki (Igels Hérissonovich Ïzhakoff-Amursky) • (yo?); June 2, 2011; 15:30 (UTC)

Thank you. Perhaps there is a technological approach that would make editing 'reference-heavy articles' easier and less error-prone (though the necessary software might be impractically difficult to create and make foolproof).

I think it is correct to say that, for reasons I and other editors have mentioned above, "solution" of the editing problem by requiring sentence by sentence references is not a workable solution. Wanderer57 (talk) 17:20, 2 June 2011 (UTC)

Thanks. I should note, however, that while it may not be a workable solution for you, Piotrus has no problem with doing it that way and neither do I. Why should we stop doing it because others seem to be incapable of doing it the same way (or simply can't be bothered to unambiguously source everything they are writing)? Mind you, I am not pushing for sentence-by-sentence referencing to be mandatory, but I'm really rather puzzled why someone else's discomfort with this should prevent me from doing something like this and to resort to sloppy wholesale referencing instead (or see someone else mangling all those refs because they don't look pretty—pshaw!). Can we at least agree to disagree and leave editing in peace those folks who actually like being 100% sure where each individual sentence comes from?—Ëzhiki (Igels Hérissonovich Ïzhakoff-Amursky) • (yo?); June 2, 2011; 18:05 (UTC)

I don't understand why it is not "a workable solution". Too much work? Not true, as I and many others show. Looks ugly? A matter of taste. I write and read academic works and books, and sure, they look different - but you know, they are not a wiki. That Wikipedia articles have their unique style is a fact, and they will look different from most other works. Due to the wiki nature, we need a way to clearly mark facts as referenced and verified. I am not aware of a better way of doing so than with footnotes. I do agree that it would be nice to develop a tool that is a bit less intrusive, and for example, a gadget to turn off all footnotes in read mode may appeal to some. But for those who want to know if something is referenced or not, and don't want to waste time checking the revisions to see if the end-of-para ref was added by the same editor who added the rest of the content (and even that may not be enough to be sure - this per this), sentence by sentence ref is the best solution we have.

While I'd like to see sentence by sentence refs obligatory to FAs, it is for community to decide on that. If the community feels it places too much burden on editors, so be it (but I recall the same arguments and difficulties when we were introducing the very basics of inline referencing...). I do think that at the very least we should encourage and recommend sentence by sentence referencing as the right thing to do, even if it is going beyond what we require.

And certainly, removing extra references is unhelpful - good-faithed and well-meaning type of vandalism, I am afraid. One could just as well remove all footnotes, or all blue links, and so on, because they do not appear in academic journals, or irritate one sense of aesthetics... --_{Piotr Konieczny aka Prokonsul Piotrus| talk} 18:31, 2 June 2011 (UTC)

Again, I disagree with you very sharply. Sentence-by-sentence refs are, usually, a bad thing. They seem to come from an attitude that everything is contentious and that WP articles are comparable to legal briefs. Making an article less legalistic and more narrative is, other things being equal, an improvement. --Trovatore (talk) 02:01, 3 June 2011 (UTC)

As I said above single sentence sourcing is simply not a real solution to the problem it claims to solve. And just because you are happy to do unnecessary extra work, it doesn't mean that other authors are nor that they should be required to do so.

What WP could use is system similar to flagged revisions, allowing other editors to sign off/confirm content and sources. But that's another discussion.--Kmhkmh (talk) 02:51, 3 June 2011 (UTC)

sentence-by sentence citations are a very bad idea. Here's another reason: many statements involve several consecutive sentences and one citation. Rjensen (talk) 02:57, 3 June 2011 (UTC)

Agree that sentence-by sentence citations are a bad idea and unnecessary. One of the arguments for them (above) is that if there's only one citation for a paragraph, you don't know if a previous editor has changed the text in front of the tag. The solution to that problem is to never, ever change text in front of a citation unless you read the source yourself and find that the article text is not supported by the source text. Referencing every word won't do any good if an editor ignores this rule. WCCasey (talk) 05:55, 3 June 2011 (UTC)

The day may come when Wikipedia finds itself with millions of Google Books citations that are broken.

Google Books is not a reliable long term site for linking citations on Wikipeida: Google makes no promise it will maintain the archive in current format; URL formats may change; books may switch from free to pay, since publishers re-publish old titles and Google removes them as free; Google may go out of business or be acquired; Google may determine books are no longer profitable and cancel or change the service.

Yet, Google is a huge resource, what is one to do? There are two solutions: Internet Archive and HathiTrust. Both of these non-profit academic-oriented sites mirror books from Google, plus have additional books they have scanned independently. In fact both these sites are larger than Google, in terms of Public Domain titles. Most people think of Google Books is the biggest/best, but it's really a poor quality also-ran whose future is uncertain. Probably the best site for long-term linking is HathiTrust because its run and maintained by the same University libraries where Google scanned the books originally. They have stated the links will be permanent and unchanging for 1000 years (or however long these Universities are in existence). If your serious about citations to scanned books that will last, it's the one to use. Green Cardamom (talk) 20:39, 4 June 2011 (UTC)

Unfortunately though, HathiTrust requires the user to register for a Michigan University login. And even when I'd got mine, it still wouldn't let me download the book I wanted even though it is in the public domain, because I am merely a member of the public and do not belong to any of the "affiliated institutions". In other words, the whole thing is even more user-hostile than Google Books. -- Alarics (talk) 18:58, 5 June 2011 (UTC)

I can access Hathi without a login, and can freely download PD books without an affiliation. Do you have any trouble viewing or downloading this book for example? I wonder if it's a US/non-US thing, or maybe an old policy since changed. Green Cardamom (talk) 06:12, 6 June 2011 (UTC)

Well free alternatives to Google Books are welcome and the issues with Google Books are known. However that doesn't change the fact that Google Books is by far the most convenient and most comprehensive thing out there for the moment (and at least the near future). Note that in most cases archive.org or HathiTrust offer no alternative, since the most common use of Google in WP is the limited preview of authoritative recent books rather than (old) PD books and the former are not available on archive.org or HathiTrust.--Kmhkmh (talk) 06:26, 6 June 2011 (UTC)

Yes the limits to Internet Archive and Hathi are only Public Domain books. It would be nice to find a way to educate or encourage editors to use IA or HT over GB for PD titles. It's like the difference between linking to Project Gutenberg versus a banner add supported site. Green Cardamom (talk) 18:02, 6 June 2011 (UTC)

Why not create a template called {{Book link}} or {{Cite link}} and tried to encourage people to link to book pages only through this template? For now, the template would be identical to {{Google books}}, except that it takes an ISBN rather than Google ID. Then, going forward, if Google books ceases to do it's job, we can change the template and fix all of the articles in one go. ---- CharlesGillingham (talk) 07:10, 6 June 2011 (UTC)

Green Cardamom asks me "Do you have any trouble viewing or downloading this book for example?" I just tried it, and it will let me look at individual pages but I am not allowed to download the book. If I try to do so I get a message reading "Full PDF available only to authenticated users from HathiTrust partner institutions." -- Alarics (talk) 08:02, 6 June 2011 (UTC)

Your right, I see that now. I guess that makes HT less desirable than Internet Archive, which is fine since IA has the largest collection of PD titles. Green Cardamom (talk) 18:02, 6 June 2011 (UTC)

CharlesGillingham makes an interesting suggestion, but the problem will be how to inform editors that the special template exists. For instance, I do quite a lot of WP editing but this is the first I have heard of a template for {{Google books}}. -- Alarics (talk) 08:02, 6 June 2011 (UTC)

Imho it doesn't really matter which template is used or what its exact name is nor whether it uses ISBN or Google ID, what matters is that the template allows a direct jump to pages in a preview mode and it is a template at all. The latter is useful for an (easy) automated mass correction later on for the case Google stops working, but should be possible with either template.--Kmhkmh (talk) 08:47, 6 June 2011 (UTC)

I generally concur with Kmhkmh here, with the exception that I would like any book template include an ISBN number as an option. Links to any online representation to a book has the potential for breakage for the foreseeable future. --Nuujinn (talk) 10:16, 6 June 2011 (UTC)

We can't encourage or force people to use templates for citations, because they're contentious for various reasons. SlimVirgin ^{TALK|CONTRIBS} 10:23, 6 June 2011 (UTC)

{{Google books}} is an external link template, not a citation template. To cite a documented hosted on Google Books, {{cite book}} is appropriate. ---— Gadget850 (Ed) ^talk 10:34, 6 June 2011 (UTC)

That's imho not quite accurate, because the (real) citation is the book and google just provides a convience link. Whether you include such a convience link into a citation template or provide it via separate template like googlebooks or even not provide one at all, is up to you or in cases of dispute up to the (pre)existing citation/reference style of the article. Similarly it is up to you whether you want to use templates for your citaions at all.--Kmhkmh (talk) 12:12, 6 June 2011 (UTC)

Exactly! It is the book that is cited, & whatever format is used should reflect this. Johnbod (talk) 12:42, 6 June 2011 (UTC)

Ok here is something to consider. Internet Archive has copied over 1 million public domain titles from Google Books, which is AFAIK most or all of the Google Book PD collection. Internet Archive has maintained Google's metadata, so it would be possible to write a script to match up the Google Book URL with the books Internet Archive URL, and then convert citations over to Internet Archive on Wikipedia. It wouldn't be the easiest programming job, but worse things have been done. I'm not suggesting this be done now, but if things ever became desperate due to Google Fail, I think the metadata exists to switch to a new provider. As for Copyright works, different story of course. Green Cardamom (talk) 18:13, 6 June 2011 (UTC)

If material can be found online, is it enough to give a link plus enough information to aid recovery after link rot, or should editors strive for references that are complete enough to pass muster in an academic setting?

For example I have been linking to military topographic maps accessible online through Perry-Castañeda Library Map Collection for the U.S. military topos or through Poehali for Soviet counterparts.

In both cases, titles and affiliations of issuing agencies have changed a few times since comprehensive topographic mapping programs came into play during World War II. I assume that academically satisfactory citation would reflect these changes, whereas something more concise and casual might ignore them and perhaps even ignore the fact that the Soviet Union is no more!

Map index numbers are the single most critical piece of information, using a row/column scheme that has remained constant and was even remarkably simiilar between the British-American and Soviet mapping programs. If you know the index number of a map and it is available online, you can probably find it without concerning yourself with the institutional details.

Nevertheless there is a certain pride of craftsmanship in getting the institutional details correctly documented. Does Wikipedia encourage this as a matter of policy, or is just getting the index number right, good enough? LADave (talk) 16:47, 13 June 2011 (UTC)

(I originally raised this at Wikipedia talk:Citation templates, but was advised that this was a more appropriate location). When citing Print-on-Demand titles, or listing them in a bibliography, who should be cited as the publisher? See, for example, the bibliographies at Stefan Stenudd, Kerri Bennett Williamson, Garry Davis and many more, which list titles as published by BookSurge Publishing. BookSurge is actually the Print-on-Demand subsidiary of Amazon.com, and these books are in effect self-published. Should the authors be listed as publisher? RolandR (talk) 07:35, 13 June 2011 (UTC)

Why wouldn't you list Booksurge as Publisher?--Kmhkmh (talk) 07:56, 13 June 2011 (UTC)

Because it isn't a publisher; it is a Print-on-Demand company. It does not commission, edit or in any other way influence the content of what it prints, and takes no responsibility for the content. As it states on its website, "BookSurge, a member of the Amazon group of companies, is a print-on-demand company offering self-publishing services for authors and on-demand book manufacturing services for both authors and publishers". By citing BookSurge as a publisher, rather than the authors themselves, we risk granting a spurious legitimacy to what are, in effect, self-published titles. It is certainly relevant, when looking at a bibliography, to note that most, if not all, of a writer's books have been self-published; citing BookSurge as the publisher obscures this fact. RolandR (talk) 08:38, 13 June 2011 (UTC)

Well I guess that boils down to what exactly you consider a "publisher". But I think the whole thing might be misleading. The question whether or when we allow self published material is a separate from the exact form of citation.

As far as the citation I goes, I'm fine with listing the authors as publisher but imho that's not necessary, since he author is listed already anyhow and there are alternative means to indicate that no editorial publisher was involved. Something like : "John Smith: My Book. Booksurge (Self-publishing) 2011" does the job as as well.

As far as the issue of allowing self published material for sourcing is concerned we have WP:SPS.--Kmhkmh (talk) 09:14, 13 June 2011 (UTC)

I concur with Kmhkmh, and would point out that there are publishing houses whose publications we do not consider reliable for other reasons, for example, those companies recycling WP articles. I would advise caution introducing the notion that because a book comes from a "print on demand" publisher it isn't reliable, as that it likely part of the future for all publishers. As Kmhkmh note, we have WP:SPS to deal with this issue, and I think that is all that is required to deal with the reliability issue. I would also suggest that listing the "publisher" as the publisher makes sense, as that helps the reader find the book if they wish to. A parallel example would be, I think, google previews, which we often include as a courtesy, but which do not contribute to the notion of reliability. --Nuujinn (talk) 10:07, 13 June 2011 (UTC)

I suggest a more nuanced policy about print on demand books that take the publishing license into consideration. I would like to cite the publishing license for print on demand publisher Lulu (I would provide a link but it is on the blacklist). Lulu will publish any document that a user uploads, which is a great service that can be used by families and businesses to get nicely formatted and printed copy. Lulu does NOT consider such content to be published by them. For Lulu to be the publisher, the book must have a Lulu ISBN. Note also that an author can get his or her own ISBN and Lulu will publish it, but in that case it is the author who is the publisher, not Lulu. Note also that when Lulu provides the ISBN and acts as the publisher, the content is reviewed and must meet their editorial guidelines. It is likely that any real self publishing enterprise will have similiar publishing agreements in place. I know that this substantially complicates the issue, but it does better reflect the reality of the issue. I also agree that no blanket policy can be adequate for this issue: a self-published work may be perfectly reliable for some purposes and must be examined case-by-case.Jarhed (talk) 19:45, 14 June 2011 (UTC)

The owner of the ISBN determines the publisher of record, which is not precisely the same thing as the publisher.

A place like Booksurge seems more like a printer than a publisher to me.

So the point of including publisher information (back in the day, when the styles were created) is that if you read my paper and I cited some intriguing sounding source, then my citation would give you everything you needed to know about how to get a copy of the source: you (or your favorite bookseller) used the "publisher" and "location" information to mail a letter to the publisher and inquire about buying the book.

In the modern world, this information is completely unnecessary if you've got an ISBN, and irrelevant if the printer isn't interested in helping you get a copy of the book. So for print on demand, I think it would be okay to simply omit the name of the vanity press/printer/service bureau. In some instances, I'd consider adding a note that it was self-published (but not when that's obvious, e.g., everybody knows that a corporate website is written and published by the business). WhatamIdoing (talk) 23:19, 14 June 2011 (UTC)

Just noting that we shouldn't be using self-published sources except in very limited circumstances, and that a lot of the print-on-demand books listed on Amazon are copies of Wikipedia articles, so caution is needed in case we're eating our own tails. Also, I definitely would include the publisher to make it easier for readers to see that it's self-published. SlimVirgin ^{TALK|CONTRIBS} 23:47, 14 June 2011 (UTC)

But when you say you would "definitely... include the publisher", do you mean, in the examples above, BookSurge, or Stefan Stenudd, Kerri Bennett Williamson and Garry Davis. My inclination is to list the self-publishers themselves, not the imprint, since they take responsibility for the content. Adding just "BookSurge", without qualification, would not make it clear to most casual readers that this is a self-publishing imprint. RolandR (talk) 11:53, 15 June 2011 (UTC)

In the section Wikipedia:Footnotes#Citation_templates and also in the reference Wikipedia:CITE, it says that cite templates are optional and that they make editing harder. I think this guidance should be deprecated by Help:Footnotes#List-defined_references, which overcomes this difficulty.Jarhed (talk) 19:30, 14 June 2011 (UTC)

...but list-defined references are also optional. While I myself prefer citation templates, I would not support imposing them on those that dislike them - they should remain optional. Nikkimaria (talk) 19:50, 14 June 2011 (UTC)

I concur with Nikkimaria. But the bulleted to-be-avoided-point "Adding citation templates to an article that already uses one of the other citation formats listed in this guideline" under WP:CITEVAR ought IMHO also state that one should avoid removing citation templates from articles, in the light of WP:Wikiquette alerts/archive101#user:Bzuk and the removal of citation templates. --Eisfbnore ^talk 20:03, 14 June 2011 (UTC)

Forgive me but I must have miscommunicated. I think they should be recommended but optional.Jarhed (talk) 21:14, 14 June 2011 (UTC)

See Holy war. It's a perennial proposal, and every time it's suggested, we hear from people who really, really, really hate them. Also, some of the most common templates are fairly inefficient, which means that articles with many citation templates are slow to load. Someone converted Pain to manually formatted citations last year, and reported that the change actually halved the page loading time. WhatamIdoing (talk) 22:23, 14 June 2011 (UTC)

In addition to slowing down load times—sometimes to the point where it's difficult to get a page to load at all—some of them use citation styles that don't exist outside Wikipedia. So any new editor arriving at the article, who tries to use one of the many styles that do exist, finds herself having to add mistakes to her manual formatting to make it consistent with the templates. SlimVirgin ^{TALK|CONTRIBS} 22:28, 14 June 2011 (UTC)

It is arguable whether list defined references are an improvement. There are several arguments against them (for example: (1) It is more difficult to finding and fixing a problem in a footnote because it's harder to find the wikitext of the footnote in the reference section. (Try fixing a random spelling error in a footnote in Artificial intelligence, for example.) (2) List defined references encourage opaque computerese footnote names, like <ref name="SPQ49rev"/> which are illegible to a new person editing the article. (Unlike {{sfn}}, or simple <ref> tages.) I personally agree they are useful when the number citations-per-paragraph gets above a certain level. (I added them to artificial intelligence myself, because, in that case, the pros outweighed the cons.)

I just want to underline that there are many disagreements about citation formats, and we have to let editors make up their own minds. ---- CharlesGillingham (talk) 06:47, 15 June 2011 (UTC)

This is something I've been thinking about a lot lately. I agree with Nikkimaria that citation templates should remain optional. Sometimes I prefer them for consistency of formatting, but I also agree with SlimVirgin that they they impose a style that's often difficult to follow when formatting manually; I greatly dislike seeing them in edit windows (unless in a shortened format); and they do add to load times. When I first started editing I was led to believe that they had to be used. I think it's important to stress somewhere they are not necessary, and that citations can be formatted manually. Truthkeeper88 (talk) 12:07, 15 June 2011 (UTC)

• Cite templates are incomplete, inconsistent, stylistically poor, increase maintenance costs, lacking in fundamental citation automation expectations, require multiple systems to use effectively, lack stylistic automatic display options, are load time expensive, encourage poor citation research and presentation behaviour on the part of editors without a deep and thorough going understanding of the purpose and use of citations. On the other hand, I'm not going to stop editors or particular article editorial communities from using them if they feel that benefits outweigh costs. Fifelfoo (talk) 12:18, 15 June 2011 (UTC)

Actually, Wikipedia:Footnotes#Citation_templates says, "Use of such templates is neither encouraged nor discouraged; see WP:CITE." WP:CITE#Citation templates and tools says, "Citation templates are used to format citations in a consistent way. The use of citation templates is neither encouraged nor discouraged. Templates may be used or removed at the discretion of individual editors, subject to agreement with other editors on the article. Because templates can be contentious, editors should not change an article with a distinctive citation format to another without gaining consensus. Where no agreement can be reached, defer to the style used by the first major contributor." Any proposed changes to that should be discussed on the talk pages associated with those project pages.(re-edited self to remove obvious and unnecessarily argumentative sentence) Wtmitchell (talk) (earlier Boracay Bill) 23:10, 15 June 2011 (UTC)

If you scroll all the way down to just say this article.

http://en.wikipedia.org/wiki/Katie_Price

You see :

^ "Katie Price: Celeb Profile". OK! Magazine. Retrieved 6 November 2010.

Now I know how to get the reference link with ^[10] but how do you put text to the link?--Jimmyson1991 (talk) 21:29, 16 June 2011 (UTC)

<ref>Write whatever you want in between these ref tags</ref> => ^[11]. Nikkimaria (talk) 21:42, 16 June 2011 (UTC)

See WP:REFBEGIN. ---— Gadget850 (Ed) ^talk 21:58, 16 June 2011 (UTC)

I am waay new here. I was surprised to see ZERO results for my Google "smilaugh" produced ZERO results on Wikipedia.

So let's get this done - please help this is my first forte.

Urban Dictionary won with this result:

. Smilaugh 1 thumb up

Mixture between a Smile and a Laugh or Smiling and Laughing at the same time.

Thats so funny im going to Smilaugh.

Person smiling and lauging. You say: Oh your Smilaughing buy smilaugh mugs & shirts sponsor this wordsmile laugh laughing smiling funny by Coal Train28 Nov 19, 2008 share this — Preceding unsigned comment added by Chipabirdee (talk • contribs) 01:22, 29 June 2011 (UTC)

Right, it's a neologism. See WP:NEO. Regards, RJH (talk) 14:41, 29 June 2011 (UTC)

I am trying to provide more references for the Sir Matt Busby Player of the Year article, but I'm having trouble with a particular one I've found. Manchester United F.C. produces fact sheets for its museum, and one of these has a complete list of winners of the Sir Matt Busby Player of the Year award. However, I'm not sure how I would go about referencing this fact sheet in the article. Any ideas? – PeeJay 10:41, 11 July 2011 (UTC)

Any old way you want. Readers have the option of believing it or not; they just consider the source. GeorgeLouis (talk) 23:31, 19 July 2011 (UTC)

I'd try using the cite journal template, filling out as many fields as possible even if the template description doesn't seem completely natural fit wise. --LauraHale (talk) 02:37, 20 July 2011 (UTC)

Since you're using citations templates, you might consider using the generic {{Citation}} template. If you've found the list online, you could also consider {{Cite web}}. WhatamIdoing (talk) 04:20, 20 July 2011 (UTC)

It's always a good idea to provide the best cites possible for sources to improve the credibility of the article. In this case, I would list it as one would any other journal or fact sheet used as a reference. Examples can be found on the web. Given said fact sheet is disseminated by the organization that presents the award, it is very reliable and seems it would be the best source possible.There is nothing civil about Civil War.Let's Talk! 06:32, 22 July 2011 (UTC)

I don't see any guidance for citing books in electronic form, such as Kindle. While I sort of own one, if a present to my spouse counts, I don't use it, although I do use the Kindle app on an iPad, I'll give some thought to what makes sense, but I hope others can weigh in on how this should be handled.--SPhilbrickT 17:47, 23 July 2011 (UTC)

{{cite ebook}} redirects to {{cite book}}, for reasons that confused me when I first noticed that. However, really there are only two major differences: format (should note it's an ebook, and probably specify the system), and (possible) replacement of page numbers with location numbers or similar. Nikkimaria (talk) 19:15, 23 July 2011 (UTC)

I used to think of hard copy as immutable and preferable to softcopy which often disappeared from my media links. But I copied a title from my newspaper some years back. An editor, verifying this (!) found the paper in "permanent" storage in subscription online and changed it. I have no doubt he did this correctly. Various soft copies change until the paper (or whoever) finally "freezes" it.

Having said that, subscriptions aren't available to everyone. On the other hand, libraries don't keep hard copies much anymore.

Should we state a preference for "permanent" softcopy over hard copy? I am particularly thinking of periodic media here. Student7 (talk) 14:56, 29 July 2011 (UTC)

I think it depends on how the softcopy is accessed. If it is a verbatim copy of the printed work, and is accessed by entering the same terms one would use to locate the printed work, then it is a moot issue; it is just as easy to cite it as if it were the printed edition, perhaps with extra information to obtain it online. Jc3s5h (talk) 15:37, 29 July 2011 (UTC)

I don't think that this will happen often enough to bother with codifying a preference. Additionally, the newspaper itself normally will have records of every version, so any of them could (in theory) be verified by a trip to the newspaper's morgue. Having said that, it's not entirely unheard of for newspapers to, ah, "correct" embarrassing things later, so if we declare a preference for their latest publicly available version, we'd occasionally have a problem. The Tulsa Tribune, for example, famously deleted the editorial they wrote endorsing a lynch mob and starting the Tulsa race riot in 1921. Their microfiche "proves" that they published a large blank space in the paper that day. WhatamIdoing (talk) 01:05, 31 July 2011 (UTC)

Some newspapers might keep records of every version (though others certainly don't), but libraries generally do not. In the Newspaper Library in London one can usually find only the final edition for any given date. If what you are looking for was carried only in the early editions, you're probably out of luck.

On the original question, this raises a whole can of worms about such issues as whether the web version is the same as the print version (it seems that increasingly it tends not to be), and other related problems which I think Wikipedia has not even begun to address!

-- Alarics (talk) 10:28, 31 July 2011 (UTC)

Yes. Two Gannett papers I normally quote, frequently changes their titles when "preserving" it for soft copy. I suppose they look for titles that have not repeated, in order to make them unique (trying to find a reason here!  :). Sometimes they change facts in later versions as they become available for their website. I have not noticed "censorship" particularly, but sometimes the material changes sufficiently to be nearly embarrassing to the editor (me), when I have quoted from the hard copy and also enclose a url which has changed but didn't check every detail when I used it.

I gather from the above, that the general feeling is that we don't need to do anything as yet (if ever). We can raise it again if it seems to be creating general problems for a number of editors. Student7 (talk) 19:12, 31 July 2011 (UTC)

Hi, I made this post at Wp:citation templates, but I haven't had a response yet, and it is quite quiet there. If you have comments please post them there for ease. Thanks. Eldumpo (talk) 15:32, 1 August 2011 (UTC)

Having been frustrated for some time about quotations that do not include any in-text indication of who is being quoted, I created (on the fly) a new inline template to alert contributors to the need for attribution. The template, which is currently used in just one article (as I said, I created it on the fly!), is Template:Whosequote. Comments -- and assistance with documentation, etc., would be appreciated. --Orlady (talk) 16:58, 25 July 2011 (UTC)

How does the purpose of this differ from {{whom?}} ? --Redrose64 (talk) 17:12, 25 July 2011 (UTC)

{{whom?}} links to Wikipedia:Neutral point of view and sticks the article in Category:All NPOV disputes. That's not the behaviour that Orlady is looking for. -- John of Reading (talk) 17:20, 25 July 2011 (UTC)

Indeed, my new template inserts "who said this?" and links to WP:Citing sources#In-text attribution.
My template also differs from Template:Bywhom, which is more similar, but is focused on vague weasel wording, and Template:Cite quote, which indicates that there is no source cited for the quotation. Instead, I am concerned with situations where a direct quotation is presented, often with a footnote, but the "speaker" behind the direct quotation is not identified. --Orlady (talk) 17:30, 25 July 2011 (UTC)

In this situation, I use inline cites to both sources, e.g. <ref>Minsky (1969), quoted in Crevier (1993, p. 67)</ref>. ---- CharlesGillingham (talk) 01:54, 31 July 2011 (UTC)

My concern isn't related so much to citing documents so much as indicating in the text whose words are being quoted. It's the difference between 'The garden is "the most magnificent display of roses in New England"^[1]' and 'The garden was described by Matthew Minsky as "the most magnificent display of roses in New England"^[1]'. --Orlady (talk) 16:42, 2 August 2011 (UTC)

This edit request has been answered. Set the |answered= parameter to no to reactivate your request.

Member of Abbots Cross Congregational Church. Qualified Soccer Coach. Popular speaker on Ulster-Scots History

2.219.82.238 (talk) 10:39, 2 August 2011 (UTC)

I think you're on the wrong page--Jac16888 ^Talk12:03, 2 August 2011 (UTC)

The vast majority of footnotes in articles are references identifying sources. In some cases, editors use footnotes to contain explanatory material. In general, when editors include both, they do not distinguish and use a single sequential numbering for each.

I quickly scanned this guideline and did not see any discussion of the distinction between these two types of footnotes, nor any guidance on how numbering should be done in case both are used. Is it here and I missed it, or is it covered in another page?

My question is motivated by Atlantis (newspaper). (As an aside, this is the editors first article!) Back on topic, the editor used the cref template with Roman numerals for explanatory footnotes, and the usual ref template for citation footnotes. What guidance should I give the editor?--SPhilbrickT 22:22, 2 August 2011 (UTC)

My !vote would be to use the usual ref template for all notes, since that is most common in scholarly works. When you click on the link to the note, it is pretty obvious which is which. But my opinion on this is completely uniformed in terms of usual practice on WP, it is a new issue for me. --Nuujinn (talk) 22:58, 2 August 2011 (UTC)

{{Cref}}, {{ref}} and similar templates are still in use because of the major issue with reference nesting; see WP:REFNEST and T22707. Arthur Rudolph uses WP:REFGROUP to separate explanatory notes and footnotes. You can also use other labels such as alpha per WP:CITELABEL. ---— Gadget850 (Ed) ^talk 23:21, 2 August 2011 (UTC)

Using references with a group parameter (see WP:REFGROUP, which gadget mentioned above) is the best method, in my opinion.
— V = IR (Talk • Contribs) 23:55, 2 August 2011 (UTC)

Thanks, I hadn't noticed, at first, that the explanatory notes had references, but that explains it.--SPhilbrickT 00:00, 3 August 2011 (UTC)

You can use grouped references for explanatory notes. Just understand that if you also use list-defined references, you can include only one nested reference. Since Atlantis has a cite in each note, they will have to be placed in-text if you convert them. ---— Gadget850 (Ed) ^talk 02:30, 3 August 2011 (UTC)

The templates {{ref label}} and {{note label}} are useful for explanatory notes that you want to keep separate from reference citations. List of municipalities in Tennessee uses {{ref label}} for this purpose. --Orlady (talk) 17:07, 4 August 2011 (UTC)

List of municipalities in Tennessee illustrates the major issue with {{ref}}— manually created links lead to duplicate HTML ids resulting in invalid HTML; see W3C markup validation for List of municipalities in Tennessee. See User:Gadget850/Comparison of Footnote3 and Cite.php Footnote. List of municipalities in Tennessee can easily be updated to the current footnote system. ---— Gadget850 (Ed) ^talk 22:31, 4 August 2011 (UTC)

WP:SAYWHEREYOUGOTIT seems to be explicit in saying that the material that is sourced is what the editor has analysed and the in-line citation should not be given as only where the source gained information from. However, there is a different interpenetration to this at by an outreach project that are transcribing material from ARKive, and they have been transcribing material from ARKive and using the the sources within ARKive as the in-line references. Contributions to advance the discussion at welcome at Wikipedia_talk:GLAM/ARKive#WP:SAYWHEREYOUGOTIT_when_writing_in-line_citations.Snowman (talk) 09:34, 29 July 2011 (UTC)

The situation is this: ARKive is donating whatever text we want (written by their in-house biology staff) about various threatened species. Wikipedia editors are therefore copying over the text, and it's doing good things for some articles. The licensing/copyright issues are apparently all being properly handled, so there's no worries there.

Some of the editors are copying over ARKive's inline citations, exactly as if the editors themselves had read those books and websites (which they haven't, both by their own admission and by the fact that some of them are dead links, so they couldn't have even if they wanted to). Snowman has proposed that (exactly like WP:SAYWHEREYOUGOTIT says) these editors acknowledge that their source is ARKive, and that ARKive's source is the other website or book or whatever.

One editor has said that this is not necessary, because ARKive (the website) is basically just another Wikipedia editor. Another has said that this is not necessary, because if he WP:Transwiki'd an article from Simple Wikipedia, then he wouldn't acknowledge that Simple was his source.

I think that we need more people involved in this discussion. WhatamIdoing (talk) 00:58, 31 July 2011 (UTC)

The situation is just as with Simple English Wikipedia, Citizendium, or any of the other numerous sources listed at Category:Attribution templates. We can directly copy free-content material from other projects into our articles, including the citations that are already in that material. Doing so does not make the other projects into references, and we don't ordinarily cite them. Re-using free content makes that content a direct part of our article, and it makes its authors into authors of (part of) our article. The sources that those authors used, which are the sources the text is based on, don't change just because the text is copied elsewhere. — Carl (CBM · talk) 02:13, 31 July 2011 (UTC)

The situation is not the same as with other Wikipedias. For non-WMF material, we simply do not have a policy that allows you to pretend that you read source X when all you did was copy the citation to source X off of source Y. WhatamIdoing (talk) 15:13, 1 August 2011 (UTC)

Of course we do, it's our fundamental mission to be a free-content project. We can copy an article in entirety from Citizendium or any other reasonable free content non-WMF project without having to double-check all the references ourselves. This does not mean we are "pretending" to read anything. We are not authoring the text, we are re-using already-written text in accordance with Wikipedia's free-content mission. The original author is the person who is responsible for checking the sources at the time the text is written. — Carl (CBM · talk) 02:19, 2 August 2011 (UTC)

I'd agree with CBM above. No matter how text ends up in an article - whether originally written for that article, for another en.wiki article, for another WMF project article, for an article on another wiki, or for another appropriately licensed work - we should include the citations the author of the text used to support it. The whole point of free licensing is that the conditions under which the author originally prepared the text are of no relevance.

When I commit an edit that includes another author's citations (e.g. when collaborating on any Wikipedia article) I'm not claiming to have read all those works. I'm simply respecting the citations other contributors offered. Retaining citations when re-using outside free content by introducing it to Wikipedia, is the same as retaining citations when re-using free content by creating a new version of an existing Wikipedia article. Christopher Parham (talk) 04:47, 2 August 2011 (UTC)

But WP:BURDEN implies that the person adding content should have verified it. Copying the refs verbatim from Citizendium (or whereever) does not demonstrate that such verification has occurred. The Citizendium (or whatever) page could be a complete hoax; should we allow that to be presented here as if it were true? Citizendium (and suchlike) are WP:TERTIARY sources and should be treated as such. --Redrose64 (talk) 16:26, 2 August 2011 (UTC)

Absolutely. If you haven't read it then you shouldn't pretend that you have by adding citations to God knows what. Malleus Fatuorum 00:29, 3 August 2011 (UTC)

The situation being described is not about adding citations. We are talking about re-using free-content text on Wikipedia. If there are already citations in the text we are copying, we can copy them along with the text, and we are still not adding any citations to the text. In that case, the person re-using content here is not pretending to have read the sources, as they are not the person who wrote them. The same thing applies if we copy text with citations from one Wikipedia article to another, by the way, as each article is licensed as a separate work for copyright purposes. — Carl (CBM · talk) 20:48, 3 August 2011 (UTC)

I have invited folks at WT:FAC to join this discussion, because some of them are far better informed than me about complicated citation issues. WhatamIdoing (talk) 01:00, 3 August 2011 (UTC)

If there were a Shoddypedia with a habit of getting things wrong, we wouldn't want to import its information and cites directly into our articles with no warning where the information comes from. So ... where do we draw the line? - Dank (push to talk) 01:48, 3 August 2011 (UTC)

If someone persisted in adding bad content, we would tell them to stop. But the thing to remember is that this is not the same as being a "reliable source" in our jargon. For example, we can't reference one Wikipedia article in another, because Wikipedia articles are not "reliable sources", but we can copy text verbatim from one article to another. The same thing holds for other free-content material. There is also a presumption of good faith: we trust editors to do the right thing, and we trust that there are enough other editors to notice mistakes. We also have to remember that Wikipedians are not required to have any actual knowledge about the topics they edit on, but things seem to go OK. — Carl (CBM · talk) 21:14, 3 August 2011 (UTC)

Wikipedia articles are based on WP:V and WP:RS; copying text from anywhere that is not a reliable source (and that includes translating from other Wikis), instead of reading the sources yourself and writing the text yourself, amounts to using non-reliable sources and trying to pass them off as reliable. We shouldn't be importing info from anywhere. We should be reading sources and writing text based on sources. Who knows how many errors are proliferated throughout Wikis by translations from poorly sourced and poorly written articles in other Wikis: this should be forbidden. Well, by policy it is, actually, but people still do it regularly and get away with it regularly, since it's not always apparent to others that they never even read the sources. Same business here. You shouldn't be citing anything imported or translated unless you've read the sources yourself. SandyGeorgia (Talk) 01:58, 3 August 2011 (UTC)

I think Wikipedia:Do not include the full text of lengthy primary sources pretty well sums this up. ---— Gadget850 (Ed) ^talk 16:37, 3 August 2011 (UTC)

Neither Simple English Wikipedia nor Citizendium would count as a primary source in any event. And we are perfectly justified in copying an article from either of those locations here, word-for-word. When we do so, we are not using their articles as sources for our own articles. Instead, we are directly using their text as our article. This is an important difference: we would never cite either of them as a source, but we can directly use their text here as if it was written by a Wikipedian. Part of our mission as a free-content project is that we can re-use free content from other sources in Wikipedia articles in this way. There is an entire category of templates to facilitate the process. — Carl (CBM · talk) 20:45, 3 August 2011 (UTC)

The transwiki (and related) processes are 100% about dealing with the legality of incorporating the material. This has nothing to do with licensing. Those processes don't say one word about whether you can list citations to possibly non-existent and definitely non-reviewed sources as if you had actually read the sources yourself. WhatamIdoing (talk) 04:47, 4 August 2011 (UTC)

The pages describing those processes don't mention it because it has always been obvious that if you are copy text from one place to another, you aren't going to go through and remove all the references that it contains. There is nothing special about references added by Wikipedians versus references added by non-Wikipedians: the person who adds the references originally should have read the sources, whether that person is a Wikipedian or not. But someone who later moves the text from one place to another is not obligated to re-read the references, because they aren't the person who wrote the text. Of course if I thought the references were false I wouldn't copy the text in the first place. But we assume good faith, which means that unless there is some specific reason to think otherwise we assume the references do exist and that the person who added them to the text actually read them. — Carl (CBM · talk) 11:27, 4 August 2011 (UTC)

To summarize— someone can create their own wiki, create content, add references and release the content under CCA. Then I can copy it wholesale into a Wikipedia article with no other checks required. ---— Gadget850 (Ed) ^talk 17:30, 4 August 2011 (UTC)

No, you can't - it has to meet our policies and guidelines. Nikkimaria (talk) 17:50, 4 August 2011 (UTC)

Let's assume for the sake of discussion that the text would be acceptable if it had been typed here first. That's what I did in my comment dated 17:51 today. — Carl (CBM · talk) 17:53, 4 August 2011 (UTC)

Of course – that's the point of us having a free copyright license. Two examples: the people at Citizendium spent a long time discussing what license to use exactly because they wanted to discuss whether we could do this with their articles. And PlanetMath chagned from GFDL to CC just about a month after we did, so the licenses are still compatible.

Re-used free content is no different from content that is first published here. It's no different than if the original author had typed that text directly into Wikipedia, except that we use templates to track the copyright dependency. In particular, any editor can put a {{cn}} tag on the newly added text, and we can't use the other wiki as a source, so we would have to either find a reliable source for the disputed statement or remove the disputed statement. But if the other wiki uses sources for their text which we consider reliable, and there's no reason to think that the other wiki is more likely to be flawed than an anonymous non-expert Wikipedian, then there's no reason to double-check all the references in order to just copy text from one location to another. — Carl (CBM · talk) 17:51, 4 August 2011 (UTC)

Licensing is not the only issue to consider, though: if you import text from elsewhere without consulting the sources used and it turns out that the material was copied from a copyrighted source, you're considered to have committed a copyright violation. It's disingenuous to add material with sources when in fact you have not consulted those sources. Nikkimaria (talk) 18:01, 4 August 2011 (UTC)

It's not disingenuous, any more than copying text from one Wikipedia article to another without consulting the sources in that text would be. It's disingenuous to be the original author and cite the sources without consulting them, but in the situations we are talking about the person editing Wikipedia is not the original author of the text, and that's OK. There is no difference in this regard between re-using material from one Wikipedia article in another Wikipedia article and re-using content from a different free-content project in a Wikipedia article. Would you say that anyone moving text from one Wikipedia article to another (say, in a merge) is responsible for verifying every reference in the text that is moved? Remember that all of our articles are independent works for the purpoes of copyright. — Carl (CBM · talk) 18:13, 4 August 2011 (UTC)

WP:CWW. In any event, unless you explicitly say that you have not verified any of the sources used, I would argue that it is most certainly disingenuous to add text from an off-wiki source, because you're representing it as text that meets our policies. YMMV. Nikkimaria (talk) 18:20, 4 August 2011 (UTC)

Right - WP:CWW does not claim that it is necessary to double-check all the references in the text you are copying, nor does it suggest that you have to make some explicit statement about that when you make the copy. In fact it doesn't mention the issue at all. But the interpretation of SAYWHEREYOUGOTIT that has been proposed here is that it would be necessary to double-check every reference when moving text to a Wikipedia different article, or when moving an article from user space to article space. That interpretation isn't correct; only the original author is responsible for reading the sources that the original author cites. The fact that Wikipedia editors are often anonymous and are not required to have any knowledge of what they write about gives us even less reason to think that text copied from another Wikipedia article is somehow more trustworthy than text copied from sites like Citizendium or PlanetMath. — Carl (CBM · talk) 18:44, 4 August 2011 (UTC)

CWW is an explicit rule that only applies when moving content within WMF projects. It says nothing about you representing the hoax I post under CC-BY-SA-released to my blog or another wiki as if you had checked my made-up sources yourself. WhatamIdoing (talk) 20:48, 4 August 2011 (UTC)

There's no more reason to think that a page on Citizendium or PlanetMath is a hoax than to think a page on Wikipedia is a hoax - in fact it seems more likely to me to find a hoax Wikipedia article than on either of those sites (or, say, a hoax in EB1911...). Any consideration about whether material is appropriate to copy into an article applies the same to Wikipedia-to-Wikipedia copies as to external-to-Wikipedia copies. In both cases we have to think about the credibility of the text, but we also have to assume good faith unless there is a particular reason not to. — Carl (CBM · talk) 11:32, 5 August 2011 (UTC)

But Wikipedia pages can contain misinformation. For example, I made major revisions and expansions to Arthur Rudolph and much of that content appears to have been copied to de:Arthur Rudolph, including the source. Every source in the English article is available online or is in my personal library. Judging from discussions on the German article, I don't think the editor who translated the content read the sources. Someone added a statement to the German article that Arthur Rudolph was an SS captain that was then copied to the English article. I raised the issue for the English article and it was removed when there was no evidence to support the statement but the information is still in the German article along with discussions.

Bottom line: I will not copy content from an open source unless and until I have personally verified the references. ---— Gadget850 (Ed) ^talk 12:56, 5 August 2011 (UTC)

In order to make Nintendo DSi#Reception more aesthetically pleasing and easier to read, I am nesting references that have 4 or more sources grouped together after a sentence. I've done this with the first sentence and would like input on what I'm doing right/wrong. « ₣M₣ » 19:34, 31 July 2011 (UTC)

Well, that's not how I'd go about it. That puts the 'citation' in with the explanatory footnotes.

The first question I'd ask is: Why do you need five sources to support that statement? WhatamIdoing (talk) 15:28, 1 August 2011 (UTC)

Generalized information like that need to be cited from multiple critics. « ₣M₣ » 14:58, 2 August 2011 (UTC)

See WP:CITEBUNDLE, but in my opinion it does not fix anything. ---— Gadget850 (Ed) ^talk 15:02, 2 August 2011 (UTC)

I don't think CITEBUNDLE will work in this instance, since four of the five are cited multiple times in the article. WhatamIdoing (talk) 17:29, 3 August 2011 (UTC)

Yeah, in this scenario it doesn't do much. Another topic, is there a template I can use for references like number 4? « ₣M₣ » 18:23, 8 August 2011 (UTC)

I take it you mean the ref constructed as <ref name=Conference2>[[#CITEREFConference2008|Satoru Iwata (2008)]], p. 2.</ref> - yes there is: {{harvnb}}, although it won't look quite the same: existing^[1] using harvnb^[2]

You also need to make sure that the |ref= in the full citation is in a format that {{harvnb}} will link to, because it won't recognise the existing |ref=CITEREFConference2008. Try |ref=harv; if that doesn't work, you'll need to manually construct it like this: |ref={{harvid|Satoru Iwata|2008}} --Redrose64 (talk) 18:46, 8 August 2011 (UTC)

An RfC is in progress at Proposal: date formats in reference sections. I have offered a counter-proposal that the "Manual of Style (dates and numbers)" defer to this guideline for dates that occur in citations. I have also proposed that this guideline be modified to recommend against all-numeric dates with the day or month first, such as writing yesterday's date as "8/7/2011", even if an external style guide calls for that format, or even if the article already uses that format. Please discuss at WT:MOSNUM. Jc3s5h (talk) 21:50, 8 August 2011 (UTC)

Which Wikipedia guideline(s) should establish citation format? Currently Citing sources, Manual of Style, and Manual of Style (date and numbers) all attempt to control citation format, and contradictions currently exist concerning date format. Jc3s5h (talk) 16:51, 13 August 2011 (UTC)

Hi Jc, you've add this to the section asking editors to refrain from changing citation styles, but it's not really about that. It's best left for the relevant MoS page:

"All-numeric date formats other than YYYY-MM-DD, because of the ambiguity concerning which number is the month and which is the day. If used, YYYY-MM-DD format should be limited to Gregorian calendar dates where the year is after 1582."

SlimVirgin ^{TALK|CONTRIBS} 08:44, 21 August 2011 (UTC)

It is somewhat related, though, because some editors change articles like , with nearly every date in yyyy-mm-dd format, to , with nearly every date in Month dd, yyyy. Editors who object generally refer to WP:DATERET and WP:CITEVAR (redirect currently not working, apparently a section title was changed recently). Gimmetoo (talk) 08:51, 21 August 2011 (UTC)

Jc's edit said that editors must not change to a numeric date format, unless it's YYY-MM-DD. He doesn't say they mustn't change to August 21, 2011, so his edit wouldn't cover your example. But whatever the rights and wrong, the example isn't about changing citation style, which is what that section is about.

It might make more sense to start a new section about that in the MoS, and to add citation style to that, rather than adding MoS issues here. SlimVirgin ^{TALK|CONTRIBS} 09:40, 21 August 2011 (UTC)

Could you clarify what you mean by "citation style", then? APA style include format specifications for dates, as do others. If a "citation style" includes not just order of author, date, title and publisher, but also punctuation and other format details of that citation, then a change to those is a change to "citation style". Gimmetoo (talk) 09:48, 21 August 2011 (UTC)

My interpretation of "citation style" is everything one expects to be consistent from one citation to another, such as date format, whether to separate fields with periods or periods, whether a number for a footnote is a superscript, in square brackets, or in parenthesis, and every other nit-picky detail you can imagine.

A recent version of MOSNUM stated "Dates in references should be uniform, and in either the prevailing date format of the article, or YYYY-MM-DD". But no article would write today's date "2011, August 21", in the body of an article, so that format would be forbidden in the citations as well. But APA style calls for that format in the citations, so MOSNUM was forbidding the use of APA style, and was in direct conflict with CITE.

CITE should be self-sufficient with respect to citation style guidance so that we can make a blanket statement in MOS and MOSNUM to refer to this guideline for citation style. The exception would be items that all published style guides that any of us have ever heard of would keep the same between article body and citations. That would include things like British vs. American spelling. Jc3s5h (talk) 10:16, 21 August 2011 (UTC)

You misunderstand: today's date in YYYY-MM-DD format wouldn't be "2011, August 21", it would be "2011-08-21". --Redrose64 (talk) 16:45, 21 August 2011 (UTC)

I was not giving today's date in YYYY-MM-DD format, I was giving today's date in the format the Publication Manual of the American Psychological Association (6th ed.) requires for publication dates. Page 185 states "For magazines, newsletters, and newspapers, give the year and the exact date of the publication (month or monthand day), separated by a comma and enclosed in parentheses (see Chapter 7, Examples 7–11)." Those examples, which appear on page 200, give the following dates: "(2008, May)", "(2008, June)", "(2006, November/December)", "(1993, September 30)", and "(2007, December 11)". Jc3s5h (talk) 17:00, 21 August 2011 (UTC)

That would imply that if we were to cite a hypothetical Chinese language source, we would be allowed to follow a hypothetical Chinese citation style that would include dates like here:

「處置失當 菲警界究責 警署署長下台 4指揮官停職」《大公報》A03，2010年8月26日

Although I wouldn't strictly have a problem with such a date format, it means gobbledegook to the vast majority of readers. Even formats such as "(2008, June)", "(2006, November/December)", "(1993, September 30)" would be at odds with MOSNUM, and be inconsistent with dates in running text. As Jc3 already notes that he expects date format such as "8/21/11" or "21/8/2011" to be overriden, thus I see MOSNUM operating as a layer of guidance supplementing it by superpositioning onto WP:CITE, overriding dates like "(2006, November/December)", "(1993, September 30)". --Ohconfucius ^¡digame! 04:04, 23 August 2011 (UTC)

APA format should be an acceptable option, no? And how would APA format be "at odds with MOSNUM"? Gimmetoo (talk) 04:57, 23 August 2011 (UTC)

Looking at another of SlimVirgin's comments, "Jc's edit said that editors must not change to a numeric date format", it says more than that. If any editor has created an article using a date format such as "8/21/11" or "21/8/2011" that notation is so ambiguous that the original editor's style decision should be overridden and a more appropriate date format used. I have never seen a published style manual that calls for any of these ambiguous formats, but if such a manual exists, it should not be used in the English Wikipedia. Jc3s5h (talk) 14:56, 21 August 2011 (UTC)

I think I'm coming to understand and agree with Jc's position on this (except I would prefer it if Wikipedia had its own single defined citation style or small set of alternative styles, so we could all get used to it). It's best if everything to do with citation formatting is dealt with on this page, and that the MoS simply say something like "however, for dates in references, see WP:CITE" (much as I think it currently does). And therefore this page should recommend against ambiguous numerical date formats in references, just as the MoS recommends against them in article bodies. --Kotniski (talk) 09:47, 22 August 2011 (UTC)

I perceive Kotniski's comment "I would prefer it if Wikipedia had its own single defined citation style or small set of alternative styles" as a matter separate from the location of citation style guidance, so I will create a new heading to discuss it. Jc3s5h (talk) 13:57, 22 August 2011 (UTC)

In the #Numbers thread Kotniski wrote "I would prefer it if Wikipedia had its own single defined citation style or small set of alternative styles". Personally, I'm not prepared to endorse citation templates, due to their bulk and performance problems. Past attempts to establish a single house style, whether citation templates or something else, did not reach consensus. Let em outline what I see as the most we could hope to achieve consensus on in the direction of standardizing styles.

• The several families of citation templates are options.
• Any style manual published by a reliable publisher, such as APA, MLA, Chicago, International Astronomical Union, or American Mathematical Society are options.
• Minor modifications to a published style manual are acceptable to take advantage of the several forms of hyperlinking.
• A reasonable number of identifying numbers, such as DOI and ISBN, may be added even though not mentioned in a published style manual.
• Editors are encouraged to create templates and/or categories and use them to label which style is in use in a particular article. This would allow us to gauge the popularity of various styles. Editors should strive to avoid uncoordinated duplication of templates or categories.

An area that would require further discussion is shortened footnotes. For articles that do not use citation templates, the only published style I know of that recognizes shortened footnotes is Chicago. Jc3s5h (talk) 13:57, 22 August 2011 (UTC)

I started documenting this at User:Gadget850/Article style options. Very incomplete. I had envisioned an editnotice that you could check the options.---— Gadget850 (Ed) ^talk 17:03, 22 August 2011 (UTC)

I've run into a curious issue. An article with >100 references had a total of 6 instances of cite templates. I rewrote those to match the other 98 or so refs. An editor reverted; I quoted CITEVAR, and the editor insisted that we had to have a discussion on the talk page and achieve consensus to rewrite any refs. If people are interpreting CITEVAR that way, then either CITEVAR needs to be rewritten, or this same principle needs to be applied to every style change, including date formats. Gimmetoo (talk) 17:59, 24 August 2011 (UTC)

I presume the editors who have contributed to the citation guideline have written academic papers at some point, and understand the essence of a citation style is consistency, at least within one paper, one article, or one chapter of a book. To believe otherwise indicates total unfamiliarity with citation in academic works. Jc3s5h (talk) 18:21, 24 August 2011 (UTC)

And that's the problem. An editor is using the CITEVAR phrasing to obstruct making the citations consistent. Gimmetoo (talk) 18:50, 24 August 2011 (UTC)

Presumably if there are mostly manual refs, they came first? If so, the templates can be removed per CITEVAR, per the "first major contributor" rule. SlimVirgin ^{TALK|CONTRIBS} 19:18, 24 August 2011 (UTC)

Then could either of you make the revert and inform the editor? I also wonder what is the appropriate censure for this sort of thing. Usually I just revert style warriors, but sometimes they keep it up. When are blocks appropriate? Gimmetoo (talk) 20:19, 24 August 2011 (UTC)

I left the two in the Further Reading section because they're consistent within the section. I have no view on whether or not they should change as well. I'm not going to inform the user as he's archived the discussion on his talk page and clearly doesn't want further posts there on this matter. DrKiernan (talk) 07:19, 25 August 2011 (UTC)

I'd like to point out that if you wish the article to attain FA status, the citation format needs to be consistent, but doesn't need to use templates. See WP:FA Criteria item 2c. --Redrose64 (talk) 11:40, 25 August 2011 (UTC)

Formatting needs to be consistent within the article. This is a dispute that needs to be resolved on the article talk page. ---— Gadget850 (Ed) ^talk 11:56, 25 August 2011 (UTC)