Wikipedia talk:Citing sources

Archives

referenced POV edits

I have noticed several users making edits which are IMHO POV but are referenced. Whenever I raise the issue of neutrality they hide behind the sources claiming that it is ok and in accordance with Wiki policy because it is referenced/cited and therefore NPOV. Is there any way round this? If that is the case it seems like madness to me since there are many sources which are not objective and clearly POV. Roydosan 09:46, 26 July 2006 (UTC)

It's OK. The way to handle referenced POV edits is to balance them with equally well-referenced statements of other important, recognized points of view. The overall balance of the section or article is important and should be discussed, of course, and the references should be such and should be made in such a way that it is easy for readers to form their own judgement of how much credibility the source has with them. Thus, for example, it is perfectly appropriate to say "J. Random Fortean, president of the Society for Promotion of the Green-Cheese-Moon Concept, says 'the moon is made of green cheese.'" The reader can decide whether they trust this person as an authority on the moon's composition.

There's no more of a problem with this than with a newspaper reporting that "Iraqi Prime Minister Nouri al-Maliki depicted Iraq as central to the war on terrorism." That doesn't meant Iraq is central to the war on terrorism, and it doesn't mean Iraq isn't central, it means the Prime Minister of Iraq expressed an opinion. Dpbsmith (talk) 16:41, 27 July 2006 (UTC)

"In Press" material

I have specific permission to cite material that will be appearing in a book from an academic publishing company later this year. Strictly speaking, such information would not be verifiable at present (since the book is not yet available), but it would be of interest to Wikipedia readers as it provides important elaboration of work already covered in Wikipedia. What is Wikipedia's policy regarding citation to work that is "in press" (i.e. accepted for publication, but not yet published)? Dragons flight 18:32, 26 July 2006 (UTC)

Don't know about Wikipedia's policy, but it was briefly discussed concerning articles in the archives: Citing stuff in press. HTH. -- Woseph 14:18, 27 July 2006 (UTC)

As the policy says "published", IMHO, you need to wait until the book is published before adding the material and citation to Wikipedia. -- Donald Albury^(Talk) 02:40, 28 July 2006 (UTC)

Question on Citing Sources for Lists

I just now created a list on wikipedia entitles List of people from Abilene, Texas. I used as references all the pages on the cities and counties in the relevant area, and the pages on the relevant individuals (for example, Stedman Graham's page lists him as an alumnus of Hardin-Simmons University). Should I attempt to find specific sources for each name mentioned or not, and how, if at all, should that be indicated? I know that this has probably been asked before, but I couldn't find the discussion. Badbilltucker 17:08, 28 July 2006 (UTC)

As far as I can tell, you didn't use any sources at all except Wikipedia articles, which are not acceptable sources. I say that you didn't use sources because you didn't cite them. Citing sources means including information in the article so that other people can find the same source you used, and either verify that you understood the source correctly, or find more information. --Gerry Ashton 18:40, 28 July 2006 (UTC)

If the Wikipedia article contains a source for the person's being from Abilene, that source should be copied into the list article. If it does not contain a source, or if it's difficult to tell whether it contains a source (e.g. no inline sources, and it's too difficult to check all the external links), then you should not put the person's name in the article. You might list it in the article's Talk page, pending further research.

I am sorry to say that when I have checked lists where the only "references" are links to Wikipedia articles, the results have been very disappointing. I've recently started checking List of bisexual people. This list tried to take a creative shortcut. It claims that if the Wikipedia article link has a plus sign next to it, that's supposed to mean that the article cites a source for the person's being bisexual. Of the first five I checked, none had a clearly identifiable reference to a source that used the word "bisexual." In one, Nina Hartley, the article used the word "bisexual" but cited no sources. One, Missy Higgins, said nothing whatsoever about her sexual orientation, and none of the general references in the article said anything about it. The other three were complicated, and I'd guess the plus sign was used in good faith, but unsatisfactory; for example, one had a citation... but it was to the person's saying "I am probably more heterosexual than I am homosexual, or even bisexual." In another, the sources appeared to support the notion that this was a gay man in a marriage of convenience to a woman, which IMHO is not at all the same thing as "bisexual."

So, no shortcuts, please. If the Wikipedia article cites a source, then copy that source into the list. If you cannot find a source, then do not put that person into the list. A Wikipedia article that asserts that someone is from Abilene without citing a source may be good grist for the mill and may suggest future entries, but should not be used to justify listing the person. In fact, just the opposite: it does justify putting a {{citation needed}} tag on the sentence that identifies the person as being from Abilene. Dpbsmith (talk) 20:28, 28 July 2006 (UTC)

No disagreement on any of the above points. However, as I think the majority of you know, the amount of work required to do any such list for a larger city, particularly, let's say, Los Angeles, would be such that no such listing would ever likely be made. For what little it might be worth, I am currently going through an old biographical dictionary which lists several people from the US, generally giving information on any locations. I shall be trying to ensure that only those people who I can specifically cite sources for as being from a location will be included on it. However, there are over four hundred such categories in the US. If any of you could help me once in a while by checking to see that no one is added to these lists without a source citation, I would be deeply grateful. Badbilltucker 13:05, 31 July 2006 (UTC)

What is the current status of "fact" tag?

I'm confused as to the current status of the {{fact}}, especially in light of this conversation. Please advise. Tewfik^Talk 21:02, 28 July 2006 (UTC)

Is there a tag for inline citations needed?

I'm running across a lot of articles that have a hefty references section (where they list all the works) but nary an inline citation. Is {{unreferenced}} really appropriate then? Is there tag we can put at the top that says "References need to be cleaned up and modified to WP:Citing sources" or something like that to get the idea across that they either have no inline citations (or very few) or they didn't format their references correctly or something? plange 03:32, 30 July 2006 (UTC)

Maybe this can be used: {{Citecheck}} ≈ jossi ≈ t • @ 03:44, 30 July 2006 (UTC)

Perfect, thanks! plange 03:50, 30 July 2006 (UTC)

I would actually prefer {{cleanup-references}}. When I tag with that for this reason, I also leave a note on the talk page, for clarity. -- Beland 02:47, 4 August 2006 (UTC)

How far does this go?

Just how far to we go with citations? Look at this: Chainmail. The article consists of literally hundreds, maybe thousands, of facts. Yet someone has picked out one of these facts (third sentence in History section) and flagged it for citation. The guy who did it gave no explanation in the edit summary as to why he thought a citation was needed. Are we saying all facts should have a citation - individually? Reading the Wiki policy I begin to wonder! I would have thought in the vast majority of articles a general citations or references are all that's needed, referring the reader to alternative, hopefully verifiable sources. Even that is not always possible. Some facts speak for themselves: Earth is the third planet out from the Sun - does that need a citation? Arcturus 22:19, 30 July 2006 (UTC)

The rule is, if an item is challenged, then it is the burden of any editor that wants to item to remain in the article to supply an appropriate citation. If a 'fact' is verifiable (published in a reliable source), then there should be no problem supplying the citation. If no editor can find a suitable reference, the item nedds to come out as unverifiable. As for the Earth being the third planet from the Sun, reliable published souces are plentiful, so what's the big problem about pointing to one? BTW, I am getting really tired of people throwing up examples like that in an effort to weaken Wikipedia policies. -- Donald Albury^(Talk) 22:33, 30 July 2006 (UTC)

"BTW, I am getting really tired of people throwing up examples like that in an effort to weaken Wikipedia policies." Are you! Well go away and find something else to do then! Earth is third planet out from the Sun (ref: the monster book of facts for five year olds). Come off it! Do you really expect people to reference facts such as those. As I say, for the most part, general references to the topic are usually sufficient. If an editor thinks that a particular fact/opinion etc. requires citation then he should go and cite it rather than passing the buck. Presumably if he knows that a citation is needed, he also knows the subject matter to a certain extent. We could all go sticking that citation tag all over the place without having to answer for it. Arcturus 22:47, 30 July 2006 (UTC)

Err, no. A request for a citation is just that: a request that the editors who added the material to the article—or else someone who merely happens to be familiar with the topic—provide a source for the fact under discussion. In the particular example you bring up (the chainmail one, not the planet one), the tag is eminently reasonable. The fact in question is not at all obvious, and without a citation, the reader has no way of determining whether it is correct. Kirill Lokshin 23:00, 30 July 2006 (UTC)

Not only is Kirill Lokshin entirely correct, but a person who wanted to learn more about the discovery of 5th century BC chainmail would have to search through all the external links in the hope that one of them might have the information. However, since the article does not have proper citations for any of its material, it is quite possible that the editor put the information in based on his/her personal recollections, and hasn't the slightest idea where the 5th century BC discovery was written about. --Gerry Ashton 23:38, 30 July 2006 (UTC)

I agree. Also, you guys might be interested in a very heated debate over inline citations at Bob Dylan plange 16:22, 1 August 2006 (UTC)

Convenience links

The text recently added is missing important giudelines. The text proposed opens a massive gap to allow the linking to purportedly reliable sources from blogs, personal websites, anonnymous websites, partisan websites, etc. The distinction needed is that the website in which the convenience link is hosted must be a reliable source on its own right. ≈ jossi ≈ t • @ 01:52, 31 July 2006 (UTC)

Yeah, good. I think so to and, actually, would prefer no personal website appear anywhere as a secondary source of information. As you know, Jossi, convenience link has been a topic of discussion at WP:RS. The "how do you make a convenience link" should go here, right? But the question, "what quality is required for a convenience link" should take place at WP:RS ? Terryeo 09:03, 31 July 2006 (UTC)

Citation Templates

I wanted to discuss the usage of various citation templates, particularly its incorporation in the form of a formal policy. I see that currently there is no strict policy on the use of citation templates. However, I am in strong favour of using them for reasons I believe are quite valid. First of all, use of citation templates ensure that the article gets uniformity of citation style and there is no difference if more than one editor edits the article (which is often the case). Secondly, they force the citations to be complete (not always, but mostly). Thirdly, their usage would allow SQL searching inside the article citations, just like Wikipedia:Persondata describes. Finally fourthly, they make the citing easy as you don't need to remember the format yourself. Other editors sometimes disagree with the fourth, but I feel it is a valid argument. Please let me know if there are other views. Note: I am not asking it to be made mandatory, though it should be recommended for the reasons described above. — Ambuj Saxena (talk) 16:09, 1 August 2006 (UTC)

I would strongly oppose this. Use of the citation templates is often too cumbersome to be useful—for articles which cite primarily large books, they're massive overkill—and forces a particular style of formatting references (MLA, if I'm not mistaken) over other equally valid ones. They should be an option to be adopted (or not) in each article as the editors of that article wish; but we should no more recommend using them than we would recommend CMS formatting over MLA formatting, or vice versa. Kirill Lokshin 16:15, 1 August 2006 (UTC)

Citation templates absolutely do not force some particular style of formatting. They simply encode the citation details in a structured way. Those can be rendered in whatever particular style guide you wish. I suppose it might be desirable for the MediaWiki developers to let editors choose the rendering on a per-article basis... but that's just technology. We need the structured data first, and in the future, such technology additions become straightforward. LotLE×talk 16:39, 1 August 2006 (UTC)

Perhaps it would be more correct to say that while the templates are theoretically metadata, the actual rendering of them in the article occurs in only one style. (And no, "Let the developers fix it" is probably not an acceptable response here; the developers have rather more important things to do than mucking around with citation formatting, I think.) Kirill Lokshin 16:57, 1 August 2006 (UTC)

I also strongly oppose recommending them. Here are some of the reasons: (a) They make texts harder to read and copy edit by making the citation longer than it needs to be. (b) They aren't flexible, and can't accommodate different needs e.g. sometimes citing works on their own in a footnote, sometimes including the citation as part of an explanation of something, or including multiple citations in one note. (c) They can't accommodate not wanting to repeat the author's name over and over, where a line ______ is used instead in the References section. (d) They take longer to write than just writing the citation out manually. (e) They force people to use a particular style. (f) In edit mode, they're harder to read than a citation written without a template. (g) They seem to add no benefit and are part of the explosion-of-template culture Wikipedia is currently experiencing. (h) If someone can work out how to write out a citation template, they can work out how to write the citation without the template. (i) They make cutting and pasting citations between articles harder, if those articles don't use the template, and most don't. SlimVirgin ^(talk) 16:28, 1 August 2006 (UTC)

Minor question: I've always seem em-dashes rather than underscores used for repeated names; is this just my not being familiar with a different formatting style? Kirill Lokshin 16:31, 1 August 2006 (UTC)

I've seen both, Kirill. SlimVirgin ^(talk) 16:33, 1 August 2006 (UTC)

Ah, ok. Kirill Lokshin 16:40, 1 August 2006 (UTC)

I'll try to find the name of the style that uses underscore. SlimVirgin ^(talk) 17:04, 1 August 2006 (UTC)

I'm with Ambuj.Saxena here. If articles move towards use of citation templates, it will eventually become possible to much more useful structured queries of citations in the WP database. Some additional tools will be needed, but they will be technically possible, rather than resemble trying to get blood from a stone (which the current unstructured data is akin to). I definitely do not think they should be suggested as mandatory, however, simply as "nice to do". I don't really get SlimVirgin's point about templates being hard to copy: they're just simply not any harder to cut-and-paste than is anything else. A given article need not use 100% citation templates to benefit from partial use of them.... but the more widely they become used, the easier a number of worthwhile things become (including, for example, creating a uniform style for citations across articles; or in some particular re-publication of WP content). LotLE×talk 16:36, 1 August 2006 (UTC)

The hard-to-copy point was that, if I'm moving citations from a page that uses templates to one that doesn't, and most don't, I have to undo the template; and in fact, it's faster just to rewrite them. I've had this happen a few times to articles I've written that have been related. I write one with lots of good citations. I write a second a few weeks later, and remember that I can use some of the same citations. I go back to the first to collect the cites, and find someone in the meantime has added the citation template. Why on earth would you add it after the fact? But some people do. And so then I can't cut and paste. SlimVirgin ^(talk) 17:08, 1 August 2006 (UTC)

Why on earth can't you cut-and-paste? I mean, I guess what you're saying is that you oppose citation templates because you oppose citation templates, right? (and therefore, you don't want to cut-and-paste them to new places). Moreover, the "inconsistent citations" matter rings hollow to my ear: pretty much the only time I ever see consistent citation formats in an article is when it actually uses templates. Every other article that theoretically chose to use some format other than what the templates currently render as is actually just a random mish-mash of semi-formatted citations. I'm sure there are exceptions out there—no doubt in articles the SlimVirgin and Kirill Lokshin have worked on—but encountering "any random article", I always find sloppy citations... and often I fix them by putting in the templates. LotLE×talk 18:19, 1 August 2006 (UTC)

You might be on to something there ;-) Perhaps the argument really boils down to this: while the templates may be suitable for cleaning up malformed citations, they are not intrinsically better than properly formatted ones, and should not be given preference in such cases. There's no reason to convert perfectly good citations to templates merely to fetishize template markup (particularly when such a conversion would result in a change of citation style within the article). Kirill Lokshin 18:28, 1 August 2006 (UTC)

Actually, partial use of citation templates can often be worse than no use at all, as identical references may be formatted very differently depending on whether or not the template is used. Kirill Lokshin 16:40, 1 August 2006 (UTC)

I guess it's a difference in time-scale of concern. I want to enable WP to be as good as possible, but looking at what can happen eventually (not decades, but quite likely months). It's true that looking towards eventual improvements might make for very slight formatting inconsistencies in the short term. But opposing citation templates essentially amounts to opposing the whole idea of a semantic web in which WP participates. I'd rather aim for best than for right now. LotLE×talk 16:43, 1 August 2006 (UTC)

Better is the enemy of good ;-) Kirill Lokshin 16:50, 1 August 2006 (UTC)

Well, the Voltaire quote is that "The best is the enemy of the good" (not the better). But I actually had in mind more of the Zen of Python:

• Now is better than never.
• Although never is often better than right now.

LotLE×talk 19:20, 1 August 2006 (UTC)

Lulu, you assume that these templates are "best"; in fact, they are not. They are one way of doing things, and a particularly cumbersome and inflexible way at that. Jayjg ^(talk) 16:52, 1 August 2006 (UTC)

No, actually the templates really are the best possible. Look beyond the surface syntax here. I don't care whether the field names happen to have upper- or lowercase, or whether squiggly brackets are used (as opposed to, say, some XML-ish angle brackets). All of this is superficial because I can write a tool that mechanically transforms the structured information in current templates into whatever hypothetical "perfect syntax" someone discovers in the future. On the other hand, neither I nor anyone else, will ever be able to reliably transform all the different styles of informal semi-random references into structured data (it happens that I quite literally "wrote the book" on this topic). For WP to really move forward in a lot of technical desiderata (e.g. structured querying of references or production of style-normalized bound versions), we have to start using templates, or at least something morally equivalent to them. LotLE×talk 18:27, 1 August 2006 (UTC)

Standard referencing would be nice, but these templates aren't the right tool for it. Jayjg ^(talk) 19:24, 1 August 2006 (UTC)

Ambuj, I strongly oppose for the reasons cogently listed by Kirill Lokshin and SlimVirgin. The first advantage you list only works if people actually use the template, and uniformity can be enforced just as easily by fixing up references, rather than pushing a template. The second advantage you list is incorrect, since the templates don't actually force any more completeness than anything else. The third advantage you list is a minor technical advantage which is of no help to article writers. The fourth advantage you list is also incorrect, since it's actually harder to remember how to use the template properly than it is to remember how to properly cite information. The template seems to be designed solely for those who like to enforce a specific style, and who like playing around with cumbersome templates; it has no advantage that I can see for actual article writers, and a number of significant disadvantages. Jayjg ^(talk) 16:52, 1 August 2006 (UTC)

Other guideline pages that mention citation templates say they are not obligatory, but are helpful. That would seem to be a suggestion. They are helpful tools for some editors. Saying (in bold) they are not recommended is discouraging their use. Gimmetrow 17:20, 1 August 2006 (UTC)

Which guideline pages are you referring to? From what I can tell the citation templates are the opposite of "helpful"; I would describe those particular templates as "fetishistic overkill". Jayjg ^(talk) 17:30, 1 August 2006 (UTC)

WP:FN in particular. Gimmetrow 17:44, 1 August 2006 (UTC)

This is the guideline for how to cite sources, not WP:FN. SlimVirgin ^(talk) 18:04, 1 August 2006 (UTC)

Your point? If the templates are to be discouraged why do they exist? Gimmetrow 18:14, 1 August 2006 (UTC)

Because template fetishists like templates. Jayjg ^(talk) 19:24, 1 August 2006 (UTC)

I would clarify some doubts. I am not saying to make them mandatory. I am asking to write it in such a way the editors are encouraged to do it. Also, I am not against multiple citation styles, as long as they are consistant within an article. Citation templates are helpful as they structure the reference. I some cases, it is difficult to see who is the author and who is the publisher. Sometimes there is confusion between the author and the name of the book. Use of citation templates can completely avoid these confusions. The way it is currently put, it discourages people to use them. I myself use citation templates as often as possible and I don't see any difficulty. Its just that I am used to it because I use them. Now the issue of notes a different one, but let me clarify here itself. I find that notes are completely different from citations and I prefer (note my words again, I am not implying a policy request) that they be kept in two different sections. The way this can be done can be seen in Rabindranath Tagore article. — Ambuj Saxena (talk) 17:48, 1 August 2006 (UTC)

Can you show me one example of where a reference is given without a template, where there is confusion between who is the author and who is the publisher, or between the name of the author and the name of the book? This guideline already says that there should be a References section as well as a Notes section. That has nothing to do with the use of templates. SlimVirgin ^(talk) 18:04, 1 August 2006 (UTC)

Here's one that I was peer reviewing where I had a hard time finding a particular author (Arifi) mentioned in the article plange 18:56, 1 August 2006 (UTC)

Take a look at Tattoo. This is an article I'm probably main editor on for the last number of months (or at least "maintaining editor"). There aren't all that many references in total, but those that there are a total mish-mash of reference styles, probably none of them matching any style guide. Except for the big block of medical-related citations, where I imposed citation templates (and Harvard referencing), and where the result looks very clean. On a side issue, those medical cites reflect an odd obsession by one editor who insisted on dumping in way more of that stuff than the article merits ("undue weight" and all). I suppose I ought to rationalize the rest of the references: if I do so, it will unquestionably be using citation templates to assure consistency. LotLE×talk 18:58, 1 August 2006 (UTC)

Why would you need a template to clean up the references? It would be easier and quicker just to do them manually. Jayjg ^(talk) 19:24, 1 August 2006 (UTC)

Wow, what a lot of discussion in only three hours!

I oppose templates because they tend to be inflexible and do not reflect the demands of the many varieties of documents that are cited. If they are to work, they need to handle all the options that are provided in the standard citing formats used in all the varied disciplines. The Chicago Style Manual (14th ed.) tried to cover all options in two chapters (pp. 487-635) and (pp. 637-699) and still has an online source to deal with additional contingencies.

Anyone proposing a template to cover all those options has a lot of reading to do if they hope to get it right. --SteveMcCluskey 19:28, 1 August 2006 (UTC)

Might as well throw my 2¢ in here. I tried using the cite templates in a few articles, but I could never remember the details, and would waste time trying to find the right template for the kind of citation I was doing, so I quickly gave it up. I think I do a decent job of formatting references, although I don't spend a lot of time worrying about which style guide I'm adhering to. In any case, my citations look better than most of the ones I see in Wikipedia. And I do get a bit annoyed when someone comes along and puts the references I've added to an article into a citation template without really improving anything, but I'm too busy adding new content to fight over them. -- Donald Albury^(Talk) 02:37, 2 August 2006 (UTC)

Readers and the future

I do not claim that citation templates particularly make matters easier for editors. But then, editors are not my main concern here: readers are. There's not any question in my mind whatsoever that the templates wind up helping readers because of the stylistic consistency they enforce (and the fact they render as an easily discernable format: I could hardly care less whether it happens to be MLA, Chicago, Oxford, whatever); most informal references are much less clear to parse, and much more often incomplete.

My real concern is with the future of Wikipedia. There are many things I would like readers to be able to do that they simply cannot do currently. Much of this boils down to the simple word "metadata". And many of these readers I am concerned about will not necessarily be reading WP via its official web site: they might be reading via some syndication of the articles, or on a CD-ROM/DVD version, or in a paper-published version, etc. Here are some questions that I believe it would be valuable to let readers answer (just examples, but they give a sense):

• Find every article the references Sigmund Freud's On Narcissism.
• Peform a "semantic proximity analysis" to extract hidden similarities and connections among referenced works (using, for example, back-propogation neural nets).
• Find out what the most widely cited book published by Oxford University Press is.
• Create an index of links (or page numbers) of all articles discussing Otto Rank.
• Print out a Wikipedia article with (automated) assurance that references match a publishers style guidelines.

Not everything "meta" about Wikipedia lies in its references, but those are a very reasonable initial target. All of these desirable future uses are "easy" once citation templates are used consistly, and impossible with the current hodge-podge of informal citation. Adding templates may not make an article much better, tomorrow; but if we start doing them, the bulk of articles can be a lot more useful next year. LotLE×talk 18:47, 1 August 2006 (UTC)

Good things "next year" (our nifty new feature development cycle seems rather longer than this, but that's somewhat beside the point) don't really justify bad things right now.

More generally, there's a confusion between metadata and the rendering of metadata that's responsible for some of these problems. Encapsulating citation information in metadata is very good. Using a raw dump of that metadata to render citations in articles is not so good, as it doesn't render very well in a great many cases. Once this (mostly technical) issue is fixed, we can start using citation templates more widely; until then, it's unreasonable to demand that editors decrease the quality of articles they're working on by introducing citations (the output of the rendering, not the metadata itself) that don't work well in context of the article. Kirill Lokshin 18:58, 1 August 2006 (UTC)

That sounds like a false dilemma to me. Can you cite one single article where the addition of citation templates decreased the quality for readers?! I've never seen that, and have trouble imagining it. I can buy that the templates ocassionally add a very slight amount of extra work for editors, but even that seems like a very mixed bag to me... I've found they also make things easier when a lot of citations share some of their details, since I can cut-and-paste those fields. Is this just about you not liking that something is bolded versus italicized, or some surface aspect like that, in the current rendering?

The problem is that we don't see the benefits of metadata until it is relatively widely used. The main gains are in the aggregation of this metadata, not in some single isolated occurrence. There are a lot of problems that can't be solved by demanding the final payout right now, before being willing to take even the smallest steps in the direction of the payout. LotLE×talk 19:10, 1 August 2006 (UTC)

I suppose some metadata is better than no metadata, but pursuading the vast majority of editors to adopt a system that only works for some citations, and requiring the editors to type regular references in situations not accomodated by the templates, may be a hard sell. One case where the existing templates don't suffice that comes to mind is giving the number and session for a US Congress document. Also, note that the templates do not say what, if any, style guide they are based on, and the description of the meaning of the various templates would probably occupy around 10 printed pages if redundancy among the templates was eliminated. In contrast, style guides devote hundreds of pages to the same topic. I suggest the existing templates are not well enough documented to pursuade many editors to use them. --Gerry Ashton 19:15, 1 August 2006 (UTC)

If there was somje kind of popup tool that will allow editors to easily use cite templates, this will be a much easier sell. As it stands, the cite formats for books and websites are more complicated than just adding a http link or an informal ref. ≈ jossi ≈ t • @ 19:21, 1 August 2006 (UTC)

I don't know about US Congress documents, but it would be great to have an appropriate template created for them... nudge, nudge Gerry Ashton. If you know the format (or really just the fields needed), please create one.

I don't plan to create a template for US Congress documents because (1) I don't know how, (2) the citation templates don't say which style guide they are following, so I wouldn't know what to aim for, and (3) when I was done, there would still be too many cases not covered by templates for me to get on the template bandwagon. --Gerry Ashton 19:57, 1 August 2006 (UTC)

I think you misunderstand what the templates are, Gerry Ashton. They don't use any style guideline, nor does it make sense for them to. What they do is hold the structured data related to a reference. How that structured data is concretely output/represented is a whole different level of concern. Just like there is bibliographic software that lets you entere structured fields, then outputs references formatted according to whatever guidelines a particular publisher requires. Style guides are an entirely orthogonal concern to templates. LotLE×talk 21:55, 1 August 2006 (UTC)

While templates may be used to store metadata (a task I would argue they are unsuited for), they are fundamentally formatting that is replicated across multiple pages. The data (which is actually stored in the template's invocation within the article) is secondary; the template itself is concerned entirely with appearance. To suggest that templates are divorced from style guides is utterly nonsensical; the entire point of having a template is not to label metadata, but to replicate a particular output based on that metadata across multiple pages. In this case, said output is a citation that matches a particular style guide. Kirill Lokshin 22:00, 1 August 2006 (UTC)

It seems to me that templates are not discussed with the same precice terminology that might be expected in a discussion of XML or C++ programming. I note, however, that Wikipedia:Template namespace begins "Templates are used to duplicate the same content across more than one page. You can change a template in one place and it will immediately propagate to the pages that use it." So I would say that the Template:cite book is a thing which exists in the Template namespace, and when invoked, it does things such as put book titles in italics, put the year of a book in parenthesis, and so on. Various style guides call for these elements to be typeset in different ways, so the template most certainly does incorporate style decisions. The invocation of a citation template in an article, in Lulu of the Lotus-Eaters' words, "hold the structured data related to a reference."

That's a clever idea, Jossi. It should be doable to write some Javascript that editors can copy to their monobook.js, that would popup a dialog box with the citation fields, and then insert a properly formatted template. Hmmm.... I promise to do that as soon as I find time :-). LotLE×talk 19:39, 1 August 2006 (UTC)

LotLE, I seem to recall you arguing strongly against what you saw as an attempt to impose footnotes on editors. If I have that right (apologies if I don't), why would you want to impose any other rigid citation system on other editors? SlimVirgin ^(talk) 19:26, 1 August 2006 (UTC)

Absolutely. I wouldn't want a guideline or policy to mandate that the citation templates must be used. And if someone (somehow) wrote an automated tool that obnoxiously spidered pages to stick in the templates, I would be outraged (*cough* ref converter *cough*). I just think this guideline (or some other guideline this points to) should have some friendly advice about why citation templates are generally a "good thing(tm)", and what future-looking advantages they have. LotLE×talk 19:39, 1 August 2006 (UTC)

LotLE, you weren't arguing against footnotes becoming mandatory; you were arguing against any hint that they were preferred or recommended, so this is looking like you want recommendations when you approve of something, and strongly oppose them when you don't. My position is consistent: we shouldn't recommend footnotes over Harvard referencing (even though I prefer footnotes), and we shouldn't recommend templates over writing citations manually. SlimVirgin ^(talk) 19:42, 1 August 2006 (UTC)

No, no, no: not at all! What I was (am) really upset about was that people were using a semi-bot to disruptively impose m:Cite.php references, contrary to explicit talk page consensus, then claiming that was justified because m:Cite.php was mandatory (or other styles were "deprecated"). In fact, I mostly use m:Cite.php myself (at least for footnotes proper), and often advocate for their use on particular pages... of course, when I do so, I go to the talk page and raise the issue; what I don't do is make a blanket change to an article I've never otherwise edited, nor intend to ever edit again. I definitely think (and always have thought) that guidelines ought to clearly state the advantages (and disadvantages) of m:Cite.php. LotLE×talk

Example of citation templates messing up editing

Look at this page in edit mode. The whole page is bad, but look particularly at the section called "Controversy about Wikipedia's origins." This is the main reason I oppose citation templates. It makes pages impossible to copy edit and almost impossible to write well. Bear in mind that this is an article with only 24 footnotes. Imagine one with 100 or more. SlimVirgin ^(talk) 19:41, 1 August 2006 (UTC)

It is funny.... those pages that editors have applied effort in providing inline citations, look very hard to edit, and those that do not have citations, well.... they need them. I am not so sure if there is am easy way to fix this, unless...

1. The text on the article is first parsed
2. Article and citations are displayed on separate HTML entities
3. Some clever AJAX enables us to add/remove/edit citations without refreshing the page

It is definitively doable, and at some point we will need this badly. Otherwise, and regardless of the format used, if we want inline citations for easying WP:V, the hard-to-edit-but-well-sourced-articles will remain as such. ≈ jossi ≈ t • @ 20:05, 1 August 2006 (UTC)

True, but I find the templates make articles even harder to edit, simply because there are more characters involved. SlimVirgin ^(talk) 20:09, 1 August 2006 (UTC)

Part of the mess is the cite.php system requiring all the ref detail inline. It would be nice if it supported having all the refs in one place, using a refname in the text, and still handled automatic numbering. But independent of the ref system, templates are a help for some editors to format citations properly; other editors don't need them. The templates in that page were used one entry per line with blank fields, and it looks like a mess. On the other hand, a template used in one line with no blank fields doesn't seem to me any worse than any other refnote, and it puts the book citation in a decent format. WP:FN says that citation templates are not obligatory, but may be helpful to get citations formatted consistently. Gimmetrow 20:16, 1 August 2006 (UTC)

Maybe a simpler way of working with this would be to have the page say that citation templates can be helpful for formatting references, but that editors who are able to do so by hand are free to avoid them. As a practical point (and perhaps the reason why this issue has come up now) is that any wording that uses "recommended" or "good idea" will cause silly objections on WP:FAC (where the articles generally have extensive citations) for articles that don't use the templates. Kirill Lokshin 20:21, 1 August 2006 (UTC)

Gimmetrow, I might be wrong, but you don't have to use the cite template inline (which I also avoid since it's so clunky). I use the cite template in the References section and then for my inline citations (which appear in Notes) I just use this (if the page number is 11 and I'm citing it more than once so I want to use the name tag: <ref name="Stevens11">Stevens, ''Name of Stevens book'', 11</ref>plange 20:44, 1 August 2006 (UTC)

The ref detail in that case is 'Stevens, Name of Stevens book, 11', and yes this lessens text clutter at the cost of you creating short notes in addition to full citations. Even with one field per line, templates are much less a clutter when collected on their own in a section. However, I would really like cite.php to allow the full citation to have the tagname "Stevens", the in-text citation to just be <ref name="Stevens" page=11/>, and to do everything else automatically. Gimmetrow 22:27, 1 August 2006 (UTC)

Yep, it would be italics and that's what I meant to show but the nowiki tag I used of course made it come out with quotes, but you knew what I meant... plange

I suspect that such a system would be utterly hellish to use for more complicated footnotes (a single footnote citing two books and containing explanatory text about the differences between the two sources, for example); at the very least, therefore, we'd need to make sure the raw <ref>Text.</ref> format still worked properly. Kirill Lokshin 22:31, 1 August 2006 (UTC)

No reason the "raw" version couldn't be kept working; it would just be a variant for the code to handle. Apparently people have written code for this already. Gimmetrow 23:56, 1 August 2006 (UTC)

Not recommended

I disagree with the recent additions of the phrases "no recommendation" and "neither recommended nor required" in relation to citation templates. One is plased in a section by itself, and another is placed in bold. This discourages the use of templates which are useful to some editors, and elsewhere suggested. I would like these additions either removed or rephrased in a way that does not discourage their use. Gimmetrow 23:55, 3 August 2006 (UTC)

While the cite templates are fine for adding a new citation, I object to converting exisitng references to the cite templates when the only result is to complicate the source page without improving the displayed page. -- Donald Albury^(Talk) 00:49, 4 August 2006 (UTC)

Gimmetrow, it's a matter of fact that, as things stand, Wikipedia neither recommends nor requires the use of citation templates. That's all this guidelines says. Most Wikipedians (in my experience) don't use them, and many object to them. SlimVirgin ^(talk) 08:55, 4 August 2006 (UTC)

Fine. But bold text saying "This is not recommended" is effectively a directive not to use them. Surely that is different than simply not saying anything? As you argue above, this page is about how to cite sources, and now the guideline basically says "don't use templates". I can imagine someone objecting to a FAC that uses templates based on this guideline if it stays as-is. Furthermore, using templates in a separate bibliography section is different than using them in an in-line citation, especially when written out one field per line. I agree the latter makes editing difficult and possibly should be discouraged. (Though ironically, that is only relevant with ref-style footnotes, and WP:FN linked from this page actually suggests using citation templates! That particular text has been stable for months, by the way.) The correct statement is, I think:

Wikipedia is indifferent to their use; some editors find them helpful, and other editors find them annoying, particularly when used inline in the text. As an arbitrary decision of style either is allowed, but editors should not change articles from one style to the other without consensus.

Would something like that be acceptable? Gimmetrow 15:15, 4 August 2006 (UTC)

Sounds fair to me. I've added it here. SlimVirgin ^(talk) 01:00, 5 August 2006 (UTC)

Almost all material from one source

In Manuel Antonio Chaves (which I wrote the majority of), everything is from a single biography of Chaves except for one or two comments. If I want to use in-line citations, how should I handle that? Having [1] after every paragraph seems a little too repetitive. What I'm currently thinking of is annotating the reference to the biography: "All information in this article is from this work unless otherwise specified," or some such. The disadvantage is that someone might add something with no in-line reference; it would then appear to be from the biography. Any thoughts? —JerryFriedman 17:26, 2 August 2006 (UTC)

If it's a printed work, there should be page numbers in every citation. You can then just create a new footnote for every page(s) cited, as here. Kirill Lokshin 17:41, 2 August 2006 (UTC)

Sooner or later, more sources will be cited. May as well get your page numbers next to all of the very specific details, potentially controversial facts, etc, now. This is what I did at Katyusha, using Harvard reference style. —Michael Z. 2006-08-10 05:22 Z

Scope of references

Now that an increasing number of articles are being referenced, I am more frequently questioning exactly what claims a given reference is supposed to be documenting. If a reference footnote is at the end of a paragraph, does it support the entire paragraph, or just the last sentence, or just the last phrase? It's important to disambiguate so that people who wish to challenge or remove fishy-sounding claims will know which are at least supposedly properly referenced, and which have just been stuck in randomly. I asked about this problem on the Village Pump, and the recommendation was to put a note in the footnote part briefly describing the claims being supported. If you're referencing a whole paragraph, probably you'll want to refer to a specific version of that paragraph, since unreferenced facts may sneak in. Does that make sense? Should it be added to this style guide? -- Beland 02:52, 4 August 2006 (UTC)

Citing a section

If you have an entire section with several pieces of information taken from a single source, rather than cite each individual piece (especially if its the only source for the whole section, which may include a list) what is the best way to source that entire paragraph?--Crossmr 01:59, 5 August 2006 (UTC)

It depends. If all the information comes from a single short article, or a few pages in a book, just put a footnote at the end of the section. On the other hand, if various pieces of information are from widely scattered pages in a book, I would put in a footnote with page number for each piece of information. For example, suppose I made a list of the ages of kings and queens that appear in Shakespeare, and just gave this citation:

Shakespere, W. (1975). The complete works of William Shakespeare. New York: Avenel Books.

That would not be good enough. --Gerry Ashton 02:11, 5 August 2006 (UTC)

Excellent, thank you.--Crossmr 02:13, 5 August 2006 (UTC)

Full name

I'd like to suggest adding something like this: "The first mention within the body of an article of a person used as a source should include the first name, the full name the person usually uses."

This will be more clear and easier to read. Maurreen 17:52, 5 August 2006 (UTC)

I agree, and preferably some information about who they are e.g. "John Smith, Professor of History at Yale University, writes that ..."; otherwise the reader won't know why we're quoting him. SlimVirgin ^(talk) 20:07, 5 August 2006 (UTC)

JA: This is not standard practice in citations, and there are several reasons why. I will get to them in a minute. Jon Awbrey 21:52, 6 August 2006 (UTC)

JA: For concreteness of discussion, here is a sample text:

The distinction between logical sign and extralogical sign is already clear in Frege and Peano, but the vagaries of translation obscure the issue of the exact terms that they used for it. The distinction is sharper in Whitehead and Russell, and Gödel is basically just gistifying "the system obtained by superimposing on the Peano axioms the logic of PM" when he writes the following: The basic signs of the system P are the following: I. Constants: "~" (not), "∨" (or), "Π" (for all), "0" (nought), "f" (the successor of), "(", ")" (brackets). ... (Gödel 1931/1992, p. 42). References Gödel, K. (1931), "On Formally Undecidable Propositions of Principia Mathematica and Related Systems", B. Meltzer (trans.), R.B. Braithwaite (intro.), Basic Books, New York, NY, 1962. Reprinted, Dover Publications, Mineola, NY, 1992.

JA: Reason 1. The purpose of a citation is to provide enough information that the sufficiently interested reader can check to see whether So-&-So really wrote Such-&-Such, as claimed in the article. It is neither to drop nor to praise a famous name that will impress the reader with it's authority.

JA: Reason 2. If the citation is well-formed, as illustrated in the text above, then the full name of the author will be found in the list of references. If the author is suspected of notability then there will be a redlink thereof, and if that suspicion is already confirmed then there will be an extant article that tells readers more than they ever wanted to know thereof. Jon Awbrey 22:32, 6 August 2006 (UTC)

1. What is "JA"?
2. Using the full name is at least common in works intended for a general audience. WP is intended for a general audience.
3. Using the relevant information within the body of the article is more clear and easier to read. Most people are not accustomed to using footnotes and the like.
4. What harm do you perceive?
5. I disagree with your reversion. I suggested this twice, first on 18 July. One person voiced agreement, one person said sometimes, neither you nor anyone else objected outright. You did not comment at all. Maurreen 00:02, 7 August 2006 (UTC)

Jon, I agree of course that when cited as part of a Havard reference, only the surname is used. In academia, the surname is also used in the body of the text even on first reference by some academics (though not by all), because everyone familiar with the subject will know who the main authors are. But we are writing for a general audience, and so we have to do what journalists do, which is to give the full name on first reference, and the surname thereafter. Otherwise how can our readers know who we mean? Yes, they can whizz down to the footnote and then look that person up on Google, but I don't see why they should have to, when I can easily say: "John Smith, chair of History, X University says ..." and refer to him thereafter as Smith. SlimVirgin ^(talk) 00:13, 7 August 2006 (UTC)

JA: I am not proposing any rule against an editor adding whatever extra information that he or she thinks might be useful or interesting to the intended reader. I am simply saying that there's no need and not much feasibility to imposing a rule like the one suggested. Jon Awbrey 04:16, 7 August 2006 (UTC)

Will you please answer my questions? Maurreen 10:23, 7 August 2006 (UTC)

JA: Response to questionaire:

1. What is "JA"?

JA: In this particular context "JA" = "Jon Awbrey".

2. Using the full name is at least common in works intended for a general audience. WP is intended for a general audience.

JA: Some of the things that WP is:

Imagine a world in which every person has free access to the sum of all human knowledge. That's what we're doing.

JA: This and other mission statements of WP imply the following:

1. WP is intended for everybody.
2. WP is an encyclopedia, that is, a reference work.
3. WP aims to provide free access to the $\sum$ of all human knowledge.
4. WP is sourced research.

JA: Just as I could not immediately assimilate everything in our Britannicas when I first started peering into them as a child, not every person will immediately be able to assimilate everything in WP when he or she first starts peering into it. We can only seek to accommodate every person in principle by providing links to material that he or she can immediately assimilate, until he or she is prepared to assimilate more. Nothing about the manifold target audience, however, allows us to evade the prevailing responsibilities of WP:NOR, WP:NPOV, and WP:VER. Therefore, sources must be properly cited at every level of presentation throughout WP. Jon Awbrey 17:52, 7 August 2006 (UTC)

JA: Despite occasional assertions about the intended audience of WP, I do not find any WP:Policy that dictates "General Audience" (GA) or even PG. What I do find is discussion like that at WP:AXS that indicates just how far WP is from any kind of consensus about reader models. Of course, every individual editor does the best that he or she can to make subjects accessible to a wider audience that they might have had heretofore, but there are limits both personal and disciplinary as to just how far this might be possible. Jon Awbrey 18:20, 7 August 2006 (UTC)

3. Using the relevant information within the body of the article is more clear and easier to read. Most people are not accustomed to using footnotes and the like.

4. What harm do you perceive?

5. I disagree with your reversion. I suggested this twice, first on 18 July. One person voiced agreement, one person said sometimes, neither you nor anyone else objected outright. You did not comment at all. Maurreen 00:02, 7 August 2006 (UTC)

JA: It being summer in the northern hemisphere, many of us northern hemispherists are on holiday/vacation for extended periods of time. I hope you weren't expecting some kind of qwiki-consensus on a major guideline page like this? Jon Awbrey 20:22, 7 August 2006 (UTC)

I've restored the edit because this isn't an academic journal, and publications outside academia use full names on first reference. SlimVirgin ^(talk) 21:21, 7 August 2006 (UTC)

Jon, why do you keep reverting? I can't see what you can have against this. It's simple common sense to tell readers who we're talking about. :-) SlimVirgin ^(talk) 21:40, 7 August 2006 (UTC)

JA: That was an accident that I self-reverted but you had already jumped on it. I'm in no particular hurry about this, but it should be clear that there's no consensus about this, and I think that making changes to a major guideline demands a clear and strong consensus, otherwise what it says on the front page is a lie. Nobody wants that, now do they? Jon Awbrey 21:50, 7 August 2006 (UTC)

I don't know what it says on the front page. I can't imagine the regulars on this page objecting to the addition. Can you say why you feel so strongly against it? SlimVirgin ^(talk) 21:58, 7 August 2006 (UTC)

JA: I am referring to the standard bolierplate that says this:

This page is a style guide for Wikipedia. The consensus of many editors formed the conventions described here. Wikipedia articles should heed these guidelines. Feel free to update this page as needed, but please use the discussion page to propose major changes.

JA: Given the number of editors in WP, 2 or 3 or 10 or 20 does not count as "many". So we need to ask the question: What is the chance that lots and lots of editors will support the proposed change in citation guidelines, and support it in the sense that they actually follow it? I can project from my expereience with the articles that I am familiar with that the chances are pretty close to nil. I can only guess that the supporters of this change are thinking of a whole different population of articles, or maybe they are just not thinking ahead. Pasting another directive on a guideline page that nobody will follow just leads to cynicism about WP:Policy in general, and I think that's a bad thing for WP. Jon Awbrey 02:20, 8 August 2006 (UTC)

I am in favour of the addition. Not only is does it add clarity to the cause but it is also courteous to inform the reader fully by not assuming that degree of prior knowledge. -- Alias Flood 23:39, 7 August 2006 (UTC)

JA: By way of gathering concrete data to assist the imagination, I tried the following experiments: (1) Drew the first 10 articles off the Random Article server, (2) looked at a dart-board sampling of Feature Articles, (3) dealt the first 10 articles off the top of my Watchlist. What I discovered to my shock and dismay is that your run-of-the-mill WP article is so poorly sourced that there weren't enough instances of "John Doe writes that …" in the whole bunch to be worth mentioning. So let me try another strategy. Go visit the article Truth, where I know from actual acquaintance that the editors there are expressly dedicated to the "Generally Unfamiliar Reader", and try to convince the folks there they should write their first mentions of each source in formulations like "René Descartes, late tutor of Queen Christina of Sweden, wrote …", and when you have exorcised your wits in doing that, report back here and maybe then we'll have some common experiences to commiserate over. Until then, e-dios, Jon Awbrey 02:40, 8 August 2006 (UTC)

I suggest a compromise, such as the following wording:

"When sources are mentioned within the body of an article, it is helpful to identify them clearly on the first mention. For example, this would mean including the first name and surname, that is, the full name the person usually uses. Even better is to include some information about the person's relevant background, such as, 'John Smith, a history professor at Yale University, writes that ...' " Maurreen 16:12, 8 August 2006 (UTC)

JA: Sure. I think it's all the difference in the world between "it is helpful to …" and "should …". Jon Awbrey 20:44, 8 August 2006 (UTC)

Inline citations not preferred style?

I recently was admonished for adding inline citations to an article, as the admonishing editor believed them not to be the "preferred style" of Wikipedia. Any validity to this assertion? --Tcatts 19:57, 8 August 2006 (UTC)

Wow, that editor is definitely wrong. Inline citations are now required for any FA article plange 20:00, 8 August 2006 (UTC)

Thanks, plange. It's good to know I'm on the right track. If anyone else has a point of view here, I'd be glad to hear it. --Tcatts 20:11, 8 August 2006 (UTC)

I'm a fairly new WP editor, and to me, the phrase inline citation is rather vague. Certainly WP style is to include information about citations within sentences or paragraphs, and not just put a list of references at the end of the article. However, there are several competing systems to accomplish that. It is preferred that if you edit an existing article, you follow whatever system already exists for that article, or if a change is necessary, get consensus for the change on the article's talk page before proceeding. --Gerry Ashton 20:13, 8 August 2006 (UTC)

Correct. There are several methods... you should also check with any WikiProjects that might have your article in their scope to make sure they also don't have a preferred method. But, as Gerry said, a big no-no is mixing inline citation styles in the same article.plange 20:47, 8 August 2006 (UTC)

Personally, I think you need footnotes. You shouldn't, for example, just leave external links inline. But you need the forward reference - just putting a list of references at the end of the page makes a big mess. I hope everyone will eventually use the <ref> style for making footnotes, which enables the software to do the right thing. -- Beland 04:10, 10 August 2006 (UTC)

Why do we allow so many different styles?

Harvard, CMS, plain HTML links, footnotes, cite.php, templates. Why can't we just decide on the style that works best for an online encyclopedia, standard or otherwise, and mandate its use in all articles? It's sloppy, jarring, confusing, and inconvenient when citation styles change drastically from one article to the next. — Omegatron 18:38, 9 August 2006 (UTC)

Because different fields of study have different styles of citation (because different styles do make sense depending on the types of material generally being cited); because we aim to include all such fields, we must necessarily make allowances for these differences. Same as US/UK spelling and other such issues, really. Kirill Lokshin 18:55, 9 August 2006 (UTC)

Examples of how different fields of study require different citation styles, please. — Omegatron 18:25, 10 August 2006 (UTC)

Okay, an obvious one:

• Harvard references are very good for the hard sciences, where most of the works being cited are scholarly papers, and hence have meaningful author/year sets.
• Conversely, Harvard references are not so good for fields such as history, where many of the things being cited do not have meaningful author/date sets (and, in particular, where citing such sets for printing of, say, anonymous ancient works would be rather misleading), or popular culture (as Harvard style does not lend itself to citing, say, films or TV programs).

Meanwhile...

• CMS footnotes are very good for fields such as historiography, where extensive annotation of particular citations is appropriate.
• Conversely, CMS footnotes are not so good for fields such as physics, where having them is overkill (and they're not at all familiar to the writers of the articles).

Which is not to say that we cannot jam just about anything into a particular style; but the resulting citations are usually cumbersome at best. I see no reason why each of the principal citation styles cannot be used where appropriate, and avoided where not. Kirill Lokshin 19:01, 10 August 2006 (UTC)

There are really only three styles of citing which associate a source with a particular statement: HTML links, Harvard, and footnotes. HTML links are an easy system which works for subjects referenced entirely from the web. Harvard is good for subjects having citations mainly to print sources, and not requiring notes for text commentary. Mixed print/web citations can be done with Harvard, but it seems more straightforward with footnotes. Footnotes also allow text commentary to be mixed with citations. For each of those systems, there may also be a bibliography. If so, then Citation Machine or templates may be used in the bibliography for format, or in the citation itself with footnote style. Just my synthesis. Gimmetrow 18:45, 10 August 2006 (UTC)

I would say four styles, really:

1. HTML links are suitable primarily for web sources, allowing no separate annotations, and with all quotations appearing in the text. Maybe useful for current events where most references are news items.
2. Citations pointing directly to a bibliography allows multiple citations and provides a list of references, but at most there can be a single annotation per source, of course. Quotations appear in the text. The references are typically numbered and listed in the order they are cited.
3. Harvard style references are like footnotes light: they allow multiple citations, each minimally annotated with a page number or range, and provide a single list of references in alphabetical order. Quotations appear in the text. Bibliography can be listed in alphabetical order.
4. Full-blown footnotes allow multiple citations per reference, and each can have its own quotation and/or long annotation in the note. They are a bit more complex for the reader to follow, providing a separate list of numbered notes and alphabetized references.

Each of these has its place, depending on the way an article is written and the depth to which it is annotated. Mandating one would be unnecessarily limiting writing style, like saying every sentence must be written in subject-verb-object order, or every paragraph must have a certain number of sentences. —Michael Z. 2006-08-11 01:47 Z

Template:fact

I propose that the section How to ask for citations be modified to discourage the overuse of Template:fact. Citing sources is good and necessary, but I think the ease of insertion of the template leads to cluttering of otherwise good articles. Often, a paragraph expressing many ideas needs only one reference, but the uninformed are likely to come along and litter every sentence with {{fact}} tags. A better approach is to bring it up on the talk page - that way, cooler, wiser heads prevail. Verifiability is a must, but that doesn’t mean that citations need to be applied with religious fervor to every last little detail. The use of Template:fact by inexperienced editors is out of control. For examples, see any of these articles: Weight loss, Duck and cover, Nation building, Walgreens, 1968 Democratic National Convention, University of Westminster, Albany High School (California), Haifa, IPod photo, State of Origin, David L. Gunn. I could go on all day, but you get the idea. The problem is getting worse each day: Template_talk:fact#Viral growth. I favor the deprecation of Template:fact, but until that day, let's at least stop actively encouraging people to plaster every inch of Wikipedia with this thing. I propose to change the section to read as follows (with thanks to User:Mzajac, from whom I have borrowed some of his great suggestions on Template_talk:fact):

How to ask for citations

If an article needs references but you are unable to find them yourself, you can tag the article with the templates {{unreferenced}} or {{Primarysources}}. {{fact}} should be used sparingly, if ever. Before applying {{fact}}, strongly consider one of the following alternatives:

• Look it up, correct or remove unverifiable information, and add the citation for anything that remains (strongly preferred).
• Post a note on the talk page.
• Move doubtful text to the talk page with a comment.
• Be kind to the environment. Don't litter articles with endless {{fact}} tags.!

In short, do a small thing to improve the article, rather than adding this template.

Please:

• Don't add this template when you are frustrated.
• Don't add this template as part of a dispute.
• Don't add this template just because you doubt a statement.
• Don't add this template to a low-traffic article and forget about it.

{{citation needed}} and {{cn}} both redirect to the {{fact}} tag and may be preferred by some editors, for the sake of clarity or brevity. All of these tags can be placed in the same place that you would place an inline reference.

All unsourced and poorly sourced negative material about living persons should be removed from articles and talk pages immediately. It should not be tagged. See Wikipedia:Biographies of living persons, which is policy.

dryguy 22:20, 9 August 2006 (UTC)

Support

1. Support as nominator. dryguy 22:20, 9 August 2006 (UTC)
2. Support: editors should communicate using the discussion page, not the body of the article. —Michael Z. 2006-08-10 05:30 Z
3. Strong Support: {{fact}} is so out of hand. I would make more of a plea here if this vote stood a chance, but plainly the time is not yet ripe. It will be, however, it will be— down the road a ways, when people see that they've been taken in by obsessive nitpickers (really, I'm talking about clinically retentive people here, like those voting Oppose) who, under cover of advocacy for academic-style strong sourcing, somehow managed to impose a burden of citation far beyond all known reference works, all the way up to "the sky is blue ". You'll see, it'll happen. JDG 22:26, 11 August 2006 (UTC)

Your example is poorly chosen, first because the sky is not always blue and therefore this is a "fact" that is not really quite true, and secondly, because as is so often the case in of things that are "common knowledge," it is very easily sourced. Rather than fussing about the "fact" tag, why not just say:

A field guide notes that "the blue sky is so commonplace that it is taken for granted".

One can go on to add:

The poet Robert Service says "while the blue sky bends above/You've got nearly all that matters" Songwriter Irving Berlin wrote of "Blue Skies smiling at me," airmen fly into the wild blue yonder. But the sky is not always blue. In the Bible, Jesus says to the Pharisees "When it is evening, ye say, It will be fair weather: for the sky is red". At twilight, salmon reds, oranges, purples, white-yellows, and many shades of blue can be seen. And songwriter Oscar Hammerstein's famously wrote of "when the sky is a bright canary yellow."<ref>Bauch, Marc. American Musical. Tectum Verlag. ISBN 382888458X. p. 42

It took me less than ten minutes to turn up the Schaeffer and Minnaert sources and another fifteen to find the rest. If something is really a commonly known fact, it is just not that hard to source. ~~~~

Oppose

1. [Obligatory comments about the poll being premature] Umm, yeah. As a general rule, any article where the number of citation requests is sufficient to produce "cluttering" is most emphatically not an "otherwise good article"; indeed, some of the examples you linked can only be described as borderline worthless, as they give absolutely no indication of where the information is coming from. Kirill Lokshin 22:36, 9 August 2006 (UTC)
2. Oppose this makes out that {{fact}} is being overrused. It is not. Strongly oppose! I use {{fact}} all the time - it is very useful! (obviously this doesn't apply to slander, which should just be removed). - Ta bu shi da yu 22:43, 9 August 2006 (UTC)

   By my count, this template is now in 21,634 articles, over 1.5% of all articles in English Wikipedia. During the last 23 days it has been added to 150 new articles per day. This template doesn't encourage the addition of sources, merely the addition of this template. —Michael Z. 2006-08-10 15:12 Z

3. Oppose. Dubious information needs to be called out as such, for the benefit of readers, but I'm not usually comfortable removing information unless I'm reasonably sure it's wrong. Powers 23:57, 9 August 2006 (UTC)

   That is misuse of this template: it is meant to request a citation, not to dis another editor's contribution. From the template's docs: "Do not use this tag in order to label text which appears doubtful or false" —Michael Z. 2006-08-10 15:12 Z

4. Oppose. Parts of this proposal contradict Wikipedia:Verifiability#Burden of evidence, which is policy and cannot be overturned or modified by any action in this forum. This guideline exists to implement policy, and must conform to policy. -- Donald Albury^(Talk) 01:46, 10 August 2006 (UTC)
5. Oppose. Jon Awbrey 05:02, 10 August 2006 (UTC)
6. Oppose. If there's a way to make the template less visually obtrusive that would be nice. There is an ongoing problem in that, as of 2006 the verifiability policy is mostly honored in the breach, so using the tag produces friction since editors tend to feel that the template is being used selectively... but my experience has been that tagging unsourced content frequently results in the content's being sourced. And, really, if we want Wikipedia to approach the standards of the verifiability policy, the only options are to tag unsourced content or remove it immediately. Dpbsmith (talk) 02:02, 10 August 2006 (UTC)

   Comment In reply to Donald Albury: None of the proposed changes contradict policy, but simply advocate using {{unreferenced}} on an article or section rather than {{fact}} on every sentence. {{fact}} would still be available when justified for use on single sentences, but action to delete or add the ref is preferred. dryguy 01:54, 10 August 2006 (UTC)

   Comment The whole section How to ask for citations in your proposal is trying to deprecate any action to ask for citations. In particular, I refer you the Burden of Evidence section in WP:V:

   Any edit lacking a source may be removed, but some editors may object if you remove material without giving people a chance to provide references. If you want to request a source for an unsourced statement, a good idea is to move it to the talk page. Alternatively, you may tag the sentence by adding the {{fact}} template, or tag the article by adding {{not verified}} or {{unsourced}}. Also in that case it may be helpful for your co-editors to leave a clarifying note on the talk page, for instance indicating which sources you already checked. You can also make the unsourced sentences invisible in the article by adding <!-- before the section you want to comment out and --> after it, until reliable sources have been provided. When using this "commenting out" technique it is usually best to leave a clarifying note on the talk page.

   This is policy. If you want to deprecate these templates, you are going to have to make your argument on Wikipedia talk:Verifiability, because you cannot void or modify the verifiability policy from here. -- Donald Albury^(Talk) 02:38, 10 August 2006 (UTC)

   Comment I don't see where it contradicts any part of WP:V, not even the part you cite above. Circumventing WP:V in any way is certainly not my intent. Maybe if you can cite the portion that you claim violates WP:V, we can discuss it. My proposed changes actually still contain many of the points you cite from WP:V.dryguy 02:49, 10 August 2006 (UTC)

   Comment I see the following:

   • Look it up, correct or remove unverifiable information, and add the citation for anything that remains (strongly preferred).
   • Post a note on the talk page.
   • Move doubtful text to the talk page with a comment.
   • Be kind to the environment. Don't litter articles with endless tags.!

   In short, do a small thing to improve the article, rather than adding this template.

   as an attempt to deprecate the use of the templates and discourage editors from seeking citations. Please note also that WP:V explicitly states that providing a source is the burden of the editor adding the information or the editor seeking to retain the information, and is not the burden of the editor requesting a citation. Your proposal is trying to reverse that. If an editor requests a citation for a fact, that editor is under no obligation find a citation. There are ways of being courteous about the process, but if you don't like seeing the templates, find the citations for them yourself, don't expect everyone else to stop using the templates. -- Donald Albury^(Talk) 03:09, 10 August 2006 (UTC)

   How on earth does the first imply the second!? Each of those suggestions is aimed at helping find citations, in lieu of littering articles with templates as a substitute for finding citations. —Michael Z. 2006-08-10 05:30 Z

   Comment You seem to be focused on my side comment that I am for deprecation of Template:fact. Please note that my proposed changes do not use the word deprecate - I'm fully aware that Template:fact is not likely to be deprecated any time soon, and that is not what I am proposing here. Also, note that the proposed change does not eliminate the option to use a {{fact}} tag. Furthermore, the primary point is to choose the use of {{unreferenced}} over the use of multiple {{fact}} over several consecutive sentences. This in no way contradicts any of WP:V. dryguy 03:20, 10 August 2006 (UTC)

   Comment My experience is that marking a section "unreferencedsect" usually has no effect, and I've even had annoyed-but-cooperative editors complain because I hadn't indicated which items needed references (in a situation where, In fact, all of them did...). Marking a section "unreferencedsect" basically just gives a not-very-helpful warning to readers, is typically perceived by inexperienced editors as a vague complaint by a malcontent... and once which can be safely ignored. Furthermore, if nothing has happened after a month or so, what do you do? Move the entire section to the Talk page? That's a drastic measure which is frequently perceived as hostile overreaction and evokes a hostile response.

   On the other hand, "fact" tags quite often lead to editors supplying references. This is especially true if the surrounding material has inline references, so the visual appearance is of superscripted numbers with an occasional stray superscripted "citations needed" tag. And if tagging the item doesn't evoke a citation, eventually moving it to the talk page does not usually provoke reversion; quite commonly the item will be sourced and reinserted days or weeks later.

   The "citation needed" tag on an individual item is a Good Thing because demands or seems to demand relatively little work on the part of an editor, and therefore is likely to evoke a constructive response rather than resistance. I know that when I see a "citation needed" tag, in an article in which I don't have a strong personal interest, my reaction is to see whether I can find a source quickly and easily. When an entire section is marked, my reaction is to think "well, I certainly don't have time to clean up this entire section."

   In short, tagging whole sections or articles as unreferenced generates more heat and less improvement than tagging individual items. Dpbsmith (talk) 10:03, 10 August 2006 (UTC)

   Agree wholeheartedly. SlimVirgin ^(talk) 00:01, 11 August 2006 (UTC)

7. Oppose — Omegatron 18:31, 10 August 2006 (UTC)
8. Oppose. SlimVirgin ^(talk) 00:01, 11 August 2006 (UTC)

Neutral

References in a second editing window, bibliography in a third?

Is there any way we could (at some point in the future) separate the references from the main text by having a separate editing window for references and just footnote markers in the text? This would unclutter the main window and make working on Wikipedia closer to the way editing and writing works in the word processor programs that most people are used to.

I would actually propose having a third window for the actual bibliographical details in some kind of standard formatting. Editing and adding a footnote would go something like this:

1. Write text, add footnote marker in main window.
2. Empty footnote magically appears in footnotes window.
3. Add reference details in bibliography window according to standard format depending on the kind of reference (book, journal article, online database etc)
4. Drag and drop individual reference from bibliography window to footnotes window. It is automatically reformatted in footnote style.
5. Add page number or other specifics and any other commentary needed in footnote.
6. Next time the same reference (book, article etc) is used in the text, one can just drag and drop it again.

Even more useful would be if the bibliography window was connected to a general bibliographic database. That way already entered items could be re-used not only in the same article but in others, and it would be possible to go there and add any additional metadata (biographical notes on author, different editions, availability of rare books etc) that aren't usually required to just reference the publication.

Theoretically I guess this system would even be able to allow people to set their favourite referencing style in Preferences and have all footnotes and bibliographies conform to that, unless this would require too much database resources.

But this is perhaps all science fiction at this point... Tupsharru 08:42, 10 August 2006 (UTC)

This week's Wikipedia Signpost has a notice about a project to bring WYSIWYG editing to MediaWiki, although there are no details. In the meantime, the <ref> system (see Cite.php) allows you to reuse a citation within an article. -- Donald Albury^(Talk) 11:42, 11 August 2006 (UTC)

The Incredible Shrinking Policy Of Verifiability

JA: 'Nuff said. Jon Awbrey 17:34, 10 August 2006 (UTC)

Not really, no. I don't think anyone knows what you're referring to here. Kirill Lokshin 17:37, 10 August 2006 (UTC)

JA: Irony, of course. Just my gistification of a discussion where so many people are so incredibly concerned with hiding the {Sad Fact}(1) that your average WP article is so poorly sourced and the {Sad Fact}(2) that your average WP editor is so godawfully resistant to being reminded of {Sad Fact}(1). Jon Awbrey 17:46, 10 August 2006 (UTC)

Nicely put. -- Donald Albury^(Talk) 11:35, 11 August 2006 (UTC)