Wikipedia talk:Citing sources/Archive 1

From Wikipedia, the free encyclopedia
Jump to: navigation, search


Support and opposition

Supporters of the "cite your sources" rule include: Janet Davis, Larry Sanger, drj (strongly), sjc, Mike Dill, Taw, Damian Yerrick, tbc (strongly!), AxelBoldt, Koyaanis Qatsi (mostly just concerned that controversial statements be attributed & not presented as "Wikipedia-opinion"), JHK (strongly), Dan Keshet (strongly), LDC(weakly),Vignaux (strongly), Steven G. Johnson, JHK; Ed Poor, because citing sources makes it easy for scholars to double-check information (especially important in growing or disputed fields of study, such as global warming); Steven G. Johnson; William M. Connolley (strongly).

Opponents include: 24 (Hypertext has different rules. Don't provide data easily discoverable on Google or Bookfinder--but yes do make the references visible in a printout--names of authors should ALL have their own wiki entries, even if totally obscure--that helps us differentiate--also use the longest form of the name to allow for future Albert Einsteins, etc.)


"intellectually honest"? What does that mean? It sounds suspiciously like an attempt to frame a debate. Suppose I call this rule "intellectually paranoid." Wouldn't that strike you as curious? (I am being serious, btw--what does it mean?) --KQ

It's typically considered intellectually dishonest to pass off another person's work as your own. "Intellectually honest" in this case means "honest about representing what one's sources are." Wikipedia is an interesting case, however, in that we don't claim most of the articles as our own, nor do we claim that the articles are original work. If, therefore, the only reason for citing sources is to make sure that no mistakenly believes that an article is the author's own work, we might as well not cite sources, because no one can labor under that misconception on Wikipedia--because there are no authors per se!

But there is value in acknowledging other creative individuals as the source of Nupedia's information. We just looked stuff up, in some cases anyway, and we did it without citing what we looked up. We aren't taking credit for this, but someone might mistakenly think that we are just really smart and we did all this research ourselves.

Maybe the better argument for citing sources is just to give people good links to further reading.  :-) --LMS

I hadn't considered the fact that we aren't claiming individual ownership of what we write. It just felt icky to me to use others' work and not acknowledge it. I also found myself consulting outside sources when I doubted some articles (and usually found the articles were, in fact, correct); providing reputable references may add strength to what we write when none of us can claim to be an expert.

Perhaps the term "intellectual honesty" doesn't need to be there, and maybe I should integrate some of what I've said here. I'll get back to it soon. -- Janet Davis

Ok. That phrase just raises hackles, sorry.  :-)

I'm not being intentionally dense (though I suspect I am being dense about it somehow, since other people seem to understand it)... but: what is the difference between looking it up in a book where someone else has already looked it up and verified it, and verifying it ourselves? At what point do we quit citing other people? (at what point is something considered well-enough known that it doesn't need to be cited?) Also, suppose I take information from the Unnameable Source, which is now in the public domain, and update it and put it here. That source lists references; am I obligated to list them as well? What about if we read something several years past but remembered it; suppose the original source of knowledge was not deduction of the facts but some long-forgotten source: are we "intellectually dishonest" if we do not cite that source? (This is not a rhetorical question, as I did read voraciously about Dave Brubeck 5 or 6 years ago, and did the same thing about Stephen King over 10 years ago: biographies, interviews, essays, prefaces, etc.)

I'm coming to believe increasingly that Thomas Jefferson should have won out in the copyright debate, as he is quite correct that once you have been given a notion, you can not rid yourself of it.... But that is not the point, as my rantings will have no effect on copyright law or the codified behavior of Intellectually Honest people. Please don't think I'm being glib; I honestly do not understand, and I don't intend the questions rhetorically. --KQ

KQ: Some of us forget stuff. Lots of stuff. All the time. For this reason, I write everything I consider professionally important into a notebook (math gets a LaTeX summary), complete with references to external material and cross-referenced internally. I try to find several different references for anything I don't understand well, because in my experience, any individual reference may be incorrect in some key point.

For wikipedia, references probably aren't very important, because encyclopedias are not viable references for scholarly work, outside of work explicitly concerning encyclopedias. (I have read at minimum several hundred papers in fields spanning geology to mathemetical mechanics and never once seen an encyclopedia reference.)

Janet: Use without attribution is definitely icky. I think its ok if one limits oneself to facts in the public domain. Example: this months Natl Geographic has an article on some island off the coast of Chile with big caves. The "public domain" part of the article could be the name of the island, the location of the island, and 1 or 2 sentences on physiography: "Has limestone, lots of rains, big caves." Anything more than that, in my opinion, better have a link back to Natl Geog.

Larry: Maybe a disclaimer on the front page, and a tiny disclaimer link on each topic page served up. At some point, some knucklehead is going to serve wikipedia with a copyright violation notice. A disclaimer might make it easier to remove offending pages without any further consequences. (Let's not talk about possible patent violations... )

One further note about encouraging outside links: it will probably encourage spam. I would just love to write up a bunch of the stuff I do professionally, then link it back to my web site!

Hope this helps. DMD

First, plagarism and copyright violation are covered elsewhere, so for the sake of discussion, let's assume those issues aren't involved.

I don't think Wikipedia articles need a lot of formal citation. An encyclopedia article, in general, presents common knowledge within a field that could lead to dozens of citations. I don't particularly feel the need to mention that I perused several books or web sites to verify that sort of thing. Perhaps the urge for formal citation comes from the academic backgrounds of many of the contributors, where the meritocracy of intellectual credit is more strongly felt than in most other areas, and is highly formalized?

If a topic within an article is obscure, controversial, or just wants emphasizing, the wiki format encourages an informal citation style ("Professor Smith, in his definitive Opus 497, indicated that blah blah ...").

An encyclopedia article usually cannot have the breadth or depth of a book or focused research effort, (although Wikipedia and similar projects may change that view), so a "Suggested Reading" or "For Additional Information" reference section may be very appropriate.

To summarize, my feeling is that citations are only appropriate where the sense of the content can be clearly attributed to a particular work, but otherwise should not be a big deal -- with many editors reviewing articles, needed citations will likely appear later if not in the original article.

Just my rambling opinion.... --loh (2001-07-05)

Sources are good for traking serious scholarship, research, history, etc. But a large number of articles here aren't that; they are simple explanations of things the reader might be unfamiliar with, but about which we know something. I think "usefulness" is the most important criterion here, rather than rigorous scholarship. What is a useful "source" to cite for a article containing one paragraph explaining that Robert Heinlein is a science fiction author and listing his stories? An article about any subject should certainly point out important works on that subject, whether or not those were the sources of the information in the article. Typically any expert's knowledge of a subject will come from dozens of sources, many of which are very good in general and many of which are too specific to be of any use to newcomers, and much of it even the expert may not remember where it came from. If I'm writing about poker strategy, it would be unthinkable not to include some mention of David Sklanky's Theory of Poker and Mike Caro's Book of Tells; but if I happen to use an example from a game I played last week, will I even remember that something obscure like that Ray Zee's Seven Card Stud High-Low for Advanced Players had a chapter on the theory behind that particular play with a similar example? If I did remember, is it useful for ordinary readers to know that if that book is otherwise not that useful to them? --LDC

Just a question about formatting. What is the proper (in Wikipedia) way to acknowledge a source in the body of the article? Do we have a way to make footnotes? Should we use the HTML href and name tags to refer to items in the bibliography at the end of the article? An example of something I've done that I'm not happy with is at the beginning of fractal.

More simply: does Wikipedia need a standard for quotations and references?

Some of the more serious articles have had numbered references in footnotes, and I don't see a problem with that. Unfortunately, the software does not yet allow page-internal links, so we can't link to them, but numbers should be fine. I don't tink we should rigidly follow CMS, though. This is a new medium, and we should explore new methods. For example, standard bibliographies don't include ISBNs. In paper books, that makes sense, because library catalogs are organized by author, topic, etc. On the web, that's monumentally stupid, because ISBNs are the key to every online database like Amazon. Let me clean up the fractal article and see if you like the style. --Lee Daniel Crocker

Yes, ISBNs must be included, they are far more important than the standard bibliographic entries which are obsolete - easily looked up given the ISBN. Same for ISSN. There are reference standards used for other online systems, check out bookfinder if you really want to find everything there is. We should be assuming that such obscure sources would need to be found somewhere like that, and it deals nicely with both current and out of print books - you could do worse than add bookfinder and google searches to every page. -24

ISBN's must be included and are the only thing needed, along with the page number, for a citation. Giving more than that just increases the chance of erroroneous citattions. Somewhere in the editing guides is a note that ISBNs don't count after a book goes out of print as if the Library of Congress didn't exist. It would be great if someone wrote a method to look up ISBNs on the LOC, making citations simple and (more) foolproof. (I know this doesn't cover 1865 Sunday School teachers' texts but it will do for modern books.) -- Rich J 22:59 15 Jun 2003 (UTC)
Author and title should be given, as ISBN's are not "the only thing needed" because they may not be unique. I have encountered at least 3 books which had the same ISBN as a different book, or at least Amazon and Half.Com thought so (perhaps they bought ISBN data from the same source). Whatever the reason for apparent ISBN duplication, we can't depend upon reality or other sites being able to uniquely identify a book based on ISBN. -- SEWilco 12:11, 4 Sep 2003 (UTC)
I guess you are right. I thought the LOC was the arbiter for ISBNs. But now that it has been privatized to Amazon we will be limited to whatever ISBNs the operator types in. Who knew that Gone With the Wind was the same book as Cauliflower Your Mouth to Health and Happiness. I suggest we use Amazon lot numbers as our single reference. -- Rich J

Write your draft, find your controversy, pick the extremes as sources, and find the foundation. -24

There are a few people here who think they are entitled to source on demand, or that their suspicion makes them experts on a topic, e.g. they can't tell that "fiat" and "military fiat" are the same thing, they can't tell that "bioregional" and "ecoregional" are the same thing, and various other failures of cognition may apply. Dictionaries may be interested in these distinctions, but if we are, it's a sign that incompetents are getting into editing. It's fair to fix and add synonyms, but there are people out there deleting whole articles because it doesn't fit their pet terminology.

24, never assume what people do or do not think or understand. Thank you. Koyaanis Qatsi
I will do so every minute of every day for the rest of my life, because it's unavoidable. We can't live examining each others' motives in infinite depth. I can cite specific examples of the behavior noted, but let's not get into it, as I didn't name names - for a reason. I think this behavior has settled down somewhat. 24

As to frequency of source quoting, there are lots of people who simply don't understand what they read and object to a line here and a fact there. Fine. Forget them. Over time, factual errors will be corrected by pedants and unless they're absolutely central to the argument (as they shouldn't be, an encyclopedia article should never be describing anything so narrowly causal as to hinge on a single fact or example) they don't affect the rest of the article.

You are assuming two things here: 1: that you do understand, and 2: that since someone disagrees with you, they must not understand. See the above. Koyaanis Qatsi
I don't see how anyone could 1. get by every day without believing that they understood the things that their lives depended on or where they had very outstanding success in the past, nor 2. mistake, as you have, a problem with a narrow range of "lots of people" (a clique), for a problem with all humanity.

Various extremely controversial topics, notably politics and religion, but in matters of the state and its use of power also economics and psychiatry, are going to necessarily require an extremely careful choice of terms. Even just to choose the terms from one side of a debate, e.g. using a word like say "sociopath" which hsa many meanings, is taking a side. Same issue as "vandal" or "miscreant" or whatever. A good example is the way a Marxist says "means of production" and a classical economics says "factors of production" - to the classical, labor is just another factor, a commodity. to most people in the world, they don't want to be treated as such with no regard to their creativity, family, social ties, etc.. That's a simple example. A more complex one arises every time you get into debates about God, e.g. many atheists will vehemently deny it's a faith even though it clearly is, the neutral position is called "agnostic" (not caring and considering both theist and atheist positions to be based on a foundation axiom that they made up).

So when making that kind of claim, on that kind of topic, attribute where you can - the debates are controversial enough you should be able to do that. But don't become afraid to write the text itself, to lay out relations between the positions of various experts, etc., there is always glue, and always trust in writing and reading. Anyone who claims otherwise and thinks they deserve to have a citation on demand is just incapable of trust and ought to be ignored.

yes, attribution is important. This, 24, is where your own writing on controversial subjects is at its weakest. And your opinion on whom I ought to trust is completely irrelevant. Koyaanis Qatsi
if you insist on citation on demand, well, there are many many long and weird articles without citations, including most of Larry's. There are some profound issues with his POV on some questions, but I don't believe that I can just bug him to tell me why he believes what he believes about the topic, whenever I want. Attribution only becomes "important" when you don't trust. Thus, people may expect more of it from me than Larry. Fine. That helps to improve the articles. They may resent this requirement of improving articles. Fine. They can do something else with their time, and others will step in. You don't seem to realize that my "own writing on controversial subjects" is *deliberately* weak on attribution. It's a *policy*. Note that there are many articles I've written that are quite over-attributed. But if the topic is controversial, why sully MY credibility with that of the sources? Fewer the better, until I know *why* the topic is controversial. Then the article can improve drastically as I focus on neutralizing places where a POV was unintentionally taken.

On controversial topics, once you know what the controversy actually is at the moment, you should be able to join forces with someone with an opposing point of view and nail down two or three sharply opposed references, and a foundation you all believe in. That obviously won't happen on the first pass, but you will identify who is actually concerned to represent the topic fairly, and who is simply insisting on their own view or whta they learned in high school as "neutral", and who is hopelessly stuck in some systemic bias. You will also find out that certain people hate you, your politics, your attitude, or your viewpoint based on things written in talk files. Fine. Let them howl.

This project cannot solve its social problems without a means of m:governance and there is seemingly zero interest in avoiding a devolution to anarchy or expanding discussion of governance beyond a clique, so forget those concerns. Push to the limit of neutrality as you understand the field in the real world, and ignore the people here who think they know what neutral means.

The correct way to find what is neutral is not by prescription but by successive refinement. In that view, sources go on at the end, not at the beginning, although if you don't have two or three trusted names in an article, you don't give your opponent anything to hinge on. So try to do that.


Your comment encouraging people not to give sources so that opponents have nothing to hinge on indicates that you are not interested in dialogue, or compromise;
ROTFL - the comment says the opposite, but in a way designed to trap people like you into revealing your biased way of reading. Obviously anyone who reads that sentence with all those negatives can interpret it as "give those trusted names" or "do not give those trusted names". you chose the latter, but I read it the former way. 24

and so indicates to me that you are here not to reach consensus, as you pretend interest in (after all, what is anarchy based on) but to push your own agenda. Mull on that if you need to. Koyaanis Qatsi

I don't, although I appreciate your answer which illustrates your assumptions. I am extremely interested in consensus process but note that it is poorly understood here. I am interested also in m:governance but I am told that m:Systemic Bias in Wikipedia does not exist or is irrelevant, and comes with a monarchy appointed long before I got here. This is all very funny, but it has nothing to do with consensus or preventing anarchy. What I need to mull on, is whether I am wasting my time educating people in consensus, why it is not unanimity, and why it is not monarchy either. I have yet to see anyone here outline what theGoverning Ontological distinction actually is, that makes them believe that a source needs to be cited, versus not. 24

24 - not on the first pass - too often, impressive-looking citations and references make people accept nonsense or heavily discredited views that were popular 200 years ago. So, find out what the controversial statements are, and only then attribute where required, simultaneously indicating who said what on the controversial matters - KQ's concern can be best dealt with by realizing we are not a community but a market... and have no "opinion"...

Citing yourself as a source?

What's the story on the external links in PUCCAMP, please? The images and info are great, although I don't like having pictures before any text, and it needs to be put into complete sentences, but I'm a little dubious of having a foreign-language link without mentioning that it is on the article page, and I'm a lot dubious about the link to the contributor's résumé -- since when are our articles signed? -- isis 01:30 Nov 8, 2002 (UTC)

I've removed the signature to the talk page, and mentioned that the site is in Portugeuse. --Camembert

Why keep it on the talk page? It's already on that user's user page, and we can get there from the history. Do we all get to put our links on the talk pages of articles now? Is that only for the new ones we start, or is that for ones we edit, too? Only major edits, or minor ones, too, like the ones I only put an image in? And am I restricted to linking it to my résumé, or can I link it to my entry in Who's Who in America, too? -- isis 07:15 Nov 8, 2002 (UTC)

Well look, if you want to remove it from the talk page, then do so, it won't bother me. And if you want to try adding links to the talk page of everything you edit, then try it and see what happens (my guess is they'll be removed if it's done en masse rather than just on one ocassion by a newcomer who doesn't know better). --Camembert

I wasn't going by the newbie's putting it there: I was going by User:Chris mahan's ratifying it and your keeping it on the 'talk' page, and now we have mav saying it's okay to have attributions on the 'talk' page. I am surprised at that (as you must be, given your prediction such postings would be removed), because I thought the 'talk' page was for discussions about the subject of the article and 'user' pages were for claiming credit for articles, but the only way I'm going to learn is by asking. -- isis 15:01 Nov 8, 2002 (UTC)

No, you can also "observe a lot by watching" as Yogi Berra might say (see Yogiisms). --Ed Poor
Normally, I would just remove the credit altogether rather than put it on talk (in fact I have done this a couple of times just now) - what can I say, I'm fickle. Personally, I probably wouldn't move article credits from the talk page (there are better things to be doing), but others might. I think there are some cases where we have copyright clearance to use something, and that goes on the talk page - such credits shouldn't be removed, of course. Otherwise, I don't think it's a very big deal - I can't speak for others. --Camembert

So, what's the home page for genocide? -- ESP 16:17 18 Jul 2003 (UTC)

Inserted into the article by Lee Daniel Crocker and removed by Stephen Gilbert:

(The Open Directory Project database would be excellent for this if it were actually open, and not under AOL's control. A wiki/FDL/public domain web directory would be an excellent complement to Wikipedia.)

Point of order: Bomis does, in fact, sponsor a free Web index known as the 3apes directory. -- NetEsq 04:42 21 Jul 2003 (UTC)

I need the help of somebody who connects through a commercial ISP or through work, not from a library or school. Can you access the URL and do you get loads of data about alcohol dehydrogenase? AxelBoldt 20:51 Mar 11, 2003 (UTC)

I'm at home - I get a login page with that URL. I tried to register but it gave me the choice of academic or commercial, I'm neither - so chose commercial and was blocked (not available to commercial users without licensing) -- sannse 21:02 Mar 11, 2003 (UTC)
Thanks, I guess we can't use that site as external reference then. AxelBoldt 21:11 Mar 11, 2003 (UTC)
There is an interface, but material is not available unless you register and pay up. -- Egil
We use books as references, and last I heard you had to pay for them. Reference away.
Plus, working on wikipedia counts as "academic" - if you twist the definitions enough... :) Martin
Ever heard of libraries?
But by all means, a reference for money is 1000 times better than no reference. -- Egil 22:05 Mar 11, 2003 (UTC)

Citation style

Surely parenthesis-author-year-parenthesis is an apropriate citation style for a technical paper, not a general encyclopedia? Popular works generally use footnotes, making the main text easier to read. See the following from Power outage (here altered slighly to follow the guidelines)

It has recently been argued on the basis of historical data (Carreras 2002a) and computer modelling (Carreras 2002b) that power grids are self-organized critical systems. These systems exhibit unavoidable (Carreras 2000) disturbances of all sizes, up to the size of the entire system, and attempts to reduce the probability of small disturbances only increase the probability of large ones (Carreras 2003). This has immediate policy implications (Carreras 2002a).

Compare with -

It has recently been argued on the basis of historical data1 and computer modelling2 that power grids are self-organized critical systems. These systems exhibit unavoidable3 disturbances of all sizes, up to the size of the entire system, and attempts to reduce the probability of small disturbances only increase the probability of large ones4. This has immediate policy implications1.

Andy G 18:10, 5 Dec 2003 (UTC)

I just finished writing entries for two separate printed technical encyclopedias, and both of them required the (Author, Date) parenthetical style. One problem with the numbered style in Wikipedia is that is really hard to maintain unless there is automatic support for it, which is not currently in place. As for it interrupting the flow of the text, I think the point is moot because encyclopedia articles tend to list a set of general reference and review articles at the end, and only occasionally need to refer to specific references in the text. See discrete Hartley transform, for example. —Steven G. Johnson 01:18, 6 Dec 2003 (UTC)
(In any case, the citation style is really the least of our worries; the vast majority of articles do not reference properly at all.)

(William M. Connolley 10:30, 6 Dec 2003 (UTC)) I like (author, date) myself, because: you can read it in the text instead of breaking up your reading by going to the footnotes; and its better for cut-and-paste.

I was trying to illustrate that (author, date) gets a bit unwieldy when you need to indicate sources for several brief facts. But I suppose what reads best is a matter of personal preference. Although as we've got a nice mechanism for on-web citations like this: [1], I feel it would be better to have a similar method for off-web citations. Maybe footnotes like this: (1), preferably with automated numbering. If it was easier to cite references maybe more people would do it. Andy G 22:55, 9 Dec 2003 (UTC)

If we want a numbered style, I agree that auto-numbering is a necessity. However, I don't like the current numbering mechanism for on-web citations—any numbering system should jump to end-notes that have more detail on the citation (and which can be printed), not a link. As far as getting more people to do it, you're attacking the wrong problem: the hard part is (a) consulting a good reference in the first place and (b) writing down the citation. The in-text pointer, whether parenthetical or numbered (if automated) is the easy part. Steven G. Johnson 03:50, 10 Dec 2003 (UTC)
I agree with Steven's assessment. Making references easier to add won't encourage users, but it will not discourage users. And it'll make it easier for those inclined to cite. I think end-notes would work best, as well. They probably represent the simplest solution to implement, too. Haphazard bracketed number links are inconsistent and confusing; I suggest that it was a mistake to make such a rudimentary citation method have such a polished final look. This page is too much of a backwater to fuel much discussion. Is their any way we can come to some kind of consensus, and show the wiki developers that a quorum have agreed on some system? Feel free to use my talk page, or propose another venue. — Johny 22:25, 16 May 2004 (UTC)

would it be too bold

..for a newcomer to replace “bulleted (*) list” with bulleted (*) list ([[Wikipedia:How_to_edit_a_page#Sections.2C_paragraphs.2C_lists_and_line| bulleted (<nowiki>*) list]]</nowiki>)

OK, it would be bold since it doesn't work.

What must be obvious to most is swimming in a sea of help references I have read in the last few hours, and a link like this would‘ve saved me the 5 min. of finding out that you type a bulleted list as you‘d like it to appear, as well as reinforce what I read on creating lists. P.S. How would I shorten the syntax for this link since it is internal to wiki, and still make it go straight to “lists and lines” P.P.S. So …“pipe “ character doesn’t work in Works WP?!? greentwin 07:28, 16 Oct 2004 (UTC)

OK...could someone take the time to show me why my link doesn't go straight to the part of the table that the TOC link I copied goes to? BTW...much kudos to the author of the fictitious book bib. entries...brilliant -Tao of Poop greentwin 07:28, 16 Oct 2004 (UTC)

Tools for bibliographic descriptions other than citing references

I started Wikipedia:WikiProject Books because the bibliographical standard for citations might be adapted to other needs. But then again it might not. AlainV 03:07, 2004 Apr 28 (UTC)


This community seems very sensitive to the issues of copyright. But when it comes to credit and citation, we do quite poorly. I'm surprised at the number of articles that refer to authors' conclusions and interpretations without crediting the author or citing the source.

Some may disagree with me, but I think this is a serious weakness. It only takes discipline to cite source; hunting down original sources takes time, effort and uncommon selflessness — let's face it, unless you happen to know the source, you're unlikely to invest much effort in uncovering the implicit citations in somebody's articles.

Integrity, transparency, credibility, honesty: that's what's at stake here. What do you think?

Making references easier to add might lower the activation energy. If I'm not aware of a system that's already in place to simplify this, please let me know. — Johny 21:48, 16 May 2004 (UTC)

I strongly agree with Johny. I am active in the area of the Catholic pages, although I contribute in other areas as well. In religion people are attached to their ideas, and so a citation is beneficial to clarify POV. My goal is always to clarify what different religions say, preferably with sources, so the reader can easily see who is saying what. Within each tradition the details are "common knowledge" but from 'outside' a source is helpful so people can make intelligent comparisons. This is very instructive and as a Catholic I have no problem whatsoever with other viewpoints being clearly presented. (Trc | [msg] 20:18, 12 Jun 2004 (UTC))
One thing I have noticed, in inviting people to present citations for their contributions, is that they will sometimes cite some small piece of their contribution. In other words, they may lodge an opinion in the article, and include in the opinion a fact; when pressed they will offer a citation for the fact but not for the opinion, thus averting the prospect of their contribution being scoped rather than being presented in a normalized way. Wikipedia is not intended as a platform for opinions in the first place; but, the 'grey area' is that sometimes it is useful that the factual existence of opinions be included in an article. It is perhaps a mark of editorial obfuscation to refuse to present a source when requested. It is also unfortunate when an editor will refuse to present a source for a contribution, because it is then impossible to verify whether they have even used the work correctly, which is theoretically possible if one can visit a library in the case of a print or subscription source. Trc | [msg] 20:51, 12 Jun 2004 (UTC)
This is a serious problem, as not citing references is often part of a violation of the NPOV rule, by making it sound like the information cited is being made by the article itself, thus making the article the source of opinions. There seems to be no awareness of this issue, and even some who oppose citations, and who want Wikipedia to look like authoritive encyclopedias where the article writer simply pronounces the truth, whether it be about fact or opinion. These people want articles to be the products of consensus, whatever the editors can agree on as "proper" going into the article, instead of the NPOV way of stating only hard facts, and attributing all opinions to some source. ChessPlayer 22:09, 16 May 2004 (UTC)
It is official policy to include references, see Wikipedia:Cite your sources. Feel free to point it out to anybody who gives you a hard time about adding references, and of course set a good example by adding them to existing articles. The hardcore approach would be just to delete text and articles that are missing proper references, heh-heh. Stan 22:33, 16 May 2004 (UTC)
I agree with you on all counts. Wikipedia does not include enough references, and Wiki markup does not make them easy to include. This is also a shortcoming of traditional encyclopedias, but not one that Wikipedia should emulate. Furthermore, when one does not have the chain of trust implicit in the traditional encyclopedia's choice of contributors, it's especially important that statements of fact be verifiable. Dpbsmith 00:27, 17 May 2004 (UTC)
The problem with citing sources is that we really are an encyclopaedia and not an original research project.

In research, you really do need to cite your sources, to prove you are enlarging on what others have created, you are developing what others have originated and not just copying what someone else has written.

In an encyclopaedia, you are referring often to what is in the public domain anyway. What is more, continuously citing others makes it so much more unreadable for the often casual reader. If you look at other encyclopaedias, how often do you come across: “(such and such) is the case, according to (author So-and-sos)?
It is different, of course, if you are actually writing about a particular concept, article, book or theory originated by another author. Of course, you must refer to every part of his work if you want to describe it.
However, take a geographical entity, when all you want to do is describe its settings, its famous buildings and scenery, it borders on the ludicrous to keep citing what other people have said about it unless their views throw a different light on what you might have seen yourself. All you need do is walk out the area concerned, clap eyes on what it is famous for and its sights and describe them.
Take an article by Tiscali Reference, an encyclopaedia, about ‘’Garcia Marquez, Gabriel (Gabo). You would look in vain for any source, where the information has come from. And why would you? It doesn’t matter whether the writer of the encyclopaedia article has read all the works of the author himself, and therefore knows them. He doesn’t have to quote himself as the source, or whether he has gleaned it from other sources. Surely, it is something that is more often than not commonly obtainable knowledge open to all of us?
After all, an encyclopaedia is a work of convenience for all of us, which if we try hard enough, we can search for laboriously ourselves, but that is what it does. It helps us find all the facts in one place and saves us time.

Dieter Simon 01:08, 17 May 2004 (UTC)

Dieter writes, "In research, you really do need to cite your sources, to prove you are enlarging on what others have created, you are developing what others have originated and not just copying what someone else has written." In an encycopedia, you need to cite your sources, to prove you are not arbitrarily enlarging on what others have created. -- Jmabel 03:42, 17 May 2004 (UTC)
Yes, Jmabel, I see your point. If you are actually quoting the words someone else has used to describe a subject, of course, you must cite this source. I was however referring to the adoption into the encyclopedia of a generally known subject, such as bibliographical details of a writer, geographical locations (often you can find this in brochures), technical details of a car engine, etc. Yes, if you are citing a specific aspect of someone else's research and findings of a subject, and you find it apposite to it, or an aspect of an article which only another source has given, then that has to be cited in detail.Dieter Simon 11:19, 17 May 2004 (UTC)

No, even for "generally known" facts a citation is a good idea. For one thing, it tells someone where they can go for more information. For another thing, even if you know something off the top of your head, searching for a good reference to cite forces you to check your facts, which might be wrong or incomplete. I could go on, but reasons for citations are already described in the article. —Steven G. Johnson 21:23, May 17, 2004 (UTC)

I'll be the first to admit that I do poorly in this area, and have to wonder where to begin to improve my referencing skills. And how does one reference what one personally knows and has observed, anyway? POV or not, one of the things that makes Wikipedia great is that it can draw on so many people's knowledge. Hmmm. ;Bear 02:16, 2004 May 18 (UTC)

Remember, the main point of references is to help the reader, not to give a map of your psyche. If you personally have knowledge of something, then that should make it all the easier to identify a good reference book/article/site for the information, both to provide a way for people to check the veracity of a Wikipedia article and to give a resource to go to for more info. Remember, you're not necessarily going to be around if someone wants clarification etc. (And when you look it up, you might surprise yourself and learn something too...often, even experts can find that they have learned only a small corner of a field.) —Steven G. Johnson 02:43, May 18, 2004 (UTC)
OK, this is all good, but how about this case in point: I can put a lot of (hopefully interesting) detail into the history of Madrid, New Mexico because I lived it. For instance, the population over the winter of 1974-75 was exactly 80, up from a low of 10 in the '60s. I did the census myself (the 80). How do I cite that? Publish it elsewhere on my own website and cite that? ;Bear 21:19, 2004 May 20 (UTC)
Remember, Wikipedia is not for original research, and your personal census counts as such. You should publish it through a normal channel, e.g. your local newspaper, and cite that. (If something is too trivial to ever appear in any print publication, then it's not encyclopedic enough for Wikipedia, for the same reason that we don't allow people to put up pages about themselves.) —Steven G. Johnson 00:09, May 21, 2004 (UTC)

My main problem with the citing of sources is that it seems to rarely be done in the body of the article. Its hard to tell if the ==References== section at the end of some articles are the actual references used to write the article, of if they are simply potential references that could be used (sort of a suggested reading). Considering we may have objections to "suggesting" reading, I propose that ==References== be potential, and that ==Sources== be actual (as I have been marking article to which I add sources). Hyacinth 00:11, 20 May 2004 (UTC)

Instead, you shouldn't add references unless you have checked them for consistency with what the article says, in which case the question of whether they were a "source" is irrelevant. (Note also that this is the standard practice for scientific publications: in many cases you might come up with an idea independently, and only later do you perform a literature search and cite the related work. Both to give credit and, again, to help the reader. But you shouldn't cite something unless you've read it, or at least know for sure what it says.) —Steven G. Johnson 00:23, May 21, 2004 (UTC)
I've repeatedly watched my in-text and end-of-article citations be deleted. It seems that no one really wants Wikipedia articles to have good documentation. It's quite discouraging, actually. —Vespristiano 13:34, 2004 May 20 (UTC)
I've never had that happen, so I suspect your case is exceptional. —Steven G. Johnson 17:33, May 20, 2004 (UTC)
Once or twice I have seen a contribution – reference pair get broken during subsequent edits. I suspect that usually it is accidental. Trc | [msg] 20:07, 12 Jun 2004 (UTC)

Web pages that are periodicals

I'd like to propose adding the following text to the policy, under the section "Web pages (not periodicals):

===Web-only periodicals===
Make the article title a link to the web page. Provide the date with as much precision as the website provides. Refer to the publication by name (Salon or Slate), not by address (,, unless, of course, the publication's name is its address (e.g., If the article is part of a column, give the name of the column (e.g. "Moneybox" above) after the title. Do not make the title of the publication a link. Be careful of linking to very new articles; it is common that an article's URL will change or disappear entirely.

Agree, or disagree? — TreyHarris 06:57, 2 Jun 2004 (UTC)

The more detailed the instructions are, the less likely people are to follow them...remember, the main difficulty is in getting people to cite in the first place. So, the main focus here should be on what information should be listed, not how. (Your recommendations above, while salutory, seem to focus more on the how; e.g. Salon can be inferred from etc. It seems unlikely that more time precision than the day of publication will be useful, although ultimately people have to apply common sense on a case-by-case basis.) I think the current simple comment about web periodicals (... for online articles, make the article title a link to the URL; it may not be possible to supply a page number in this case) is sufficient and has the advantage of brevity. —Steven G. Johnson 01:19, Jun 3, 2004 (UTC)
(If we want to be very picky about reference-formatting styles, we should be using BibTeX or the like, anyway.)
The more detailed the instructions are, the less likely people are to follow them...
Do you have a source for that statement? I would think the opposite would be true. — TreyHarris 02:15, 3 Jun 2004 (UTC)
Just common sense and experience, although I'm sure references could be found in the usability community. People find reams of style guides, policies, and instructions off-putting, and aren't going to read them, much less remember them or follow them. You want to make citations as easy as possible, since it's hard enough getting people to do them already. —Steven G. Johnson 02:24, Jun 3, 2004 (UTC)
By the way, if your response is along the lines of, but what if people are confused because they don't know the precise way to cite XYZ?, then the answer is to make it clearer that they should just enter something reasonable and not worry about whether to list Salon or —Steven G. Johnson 02:33, Jun 3, 2004 (UTC)

References at the end?

(William M. Connolley 11:14, 12 Jun 2004 (UTC)) I see the page says, put your references at the end. Even for citing web pages? This is unnatural to the flow of the page. If I write:

Stoats are very nice (square bracket)]

am I breaking the style? What would fit standard journal practice is for the text above to generate a ref [1] in the text (as it does at the moment) but also automatically puts "# [1]" in a section headed "auto generated refs" at the bottom of the page.

At the end please. That's how you know you're writing an encyclopedia entry and not a journal article. :-) But seriously, a journal article has to defend itself line-by-line, since its goal is to add to the sum of human knowledge, while the encyclopedia is just repeating what people already know.
(William M. Connolley 14:02, 12 Jun 2004 (UTC)) I don't know what pages you're writing, but thats not true of, say, Global warming.
I'm sure people do that, but it's against policy; see point 10 of Wikipedia:What_Wikipedia_is_not. Perhaps I need to go over to global warming and start deleting the stuff that's not already written up in existing publications? Stan 19:06, 12 Jun 2004 (UTC)
Please realize that, occasionally, in-text pointers to the reference are helpful even in an encyclopedia article. This is true when there are a number of references and it's not clear which reference to look up for more information on a specific item of information that the user might be interested in. I agree that, as a general rule, it's better to list a small number of comprehensive textbook/review-style references at the end so that specific pointers aren't needed, but sometimes it's unavoidable, especially in newer topics. (I've written articles for print encyclopedias, and the same guidelines applied there.) —Steven G. Johnson 20:03, Jun 12, 2004 (UTC)
I don't think print encyclopedia experience is a particularly good guide here; print works will tend to encourage longer articles and less cross-linking, so as to minimize the page-flipping, and when you do that, you're right, the links at the end won't be well-associated with relevant points in the text. But I take that as evidence that a long article needs to be broken up for the sake of the poor reader. For instance, newer topics should be split out so the back-and-forth can be addressed better, and editors can fight over them without trashing main articles that stick to the settled information. Stan 21:50, 12 Jun 2004 (UTC)
Actually, I would argue that, in Wikipedia, citations are far more important than in a print encyclopedia, because in a print encyclopedia there are editors and a company who stand behind the content and who give some assurance that the article authors are experts in a topic. Obviously, you have to use your judgement in particular cases, but I don't understand why you are digging in your heels and insisting that in-text references are never helpful or appropriate. —Steven G. Johnson 22:15, Jun 12, 2004 (UTC)
I wouldn't go so far as to say never; if you have an article that's gotten overly long and haven't had a chance to reorganize it into multiple articles, and there's a technical point with a single source in the middle, then an in-text citation would make sense. I just don't want to open the door to putting a citation in every sentence or every clause, as for instance the ridiculous example currently at the top of this talk page. I think for some people there's a great temptation to try and make an article look more authoritative by sprinkling literature citations throughout (99.9% of readers will not have access to the literature being cited, so it certainly doesn't help them in any way), but if the in-text refs really are necessary and not just for showing off, I think that bespeaks fundamental flaws in the overall article design and content. So I'd say I'm in favor of allowing but deprecating the practice. Stan 05:58, 13 Jun 2004 (UTC)
Not only that, but in a print product a clear POV is known from the outside, which fact is valuable interpretive information. Trc | [msg] 22:22, 12 Jun 2004 (UTC)
The county library where I live does not charge for interlibrary loan, so 100% of Lewis and Clark County residents have access to the large majority of references available at no charge, especially the ones I have added, as they were through ILL at that library. Hyacinth 09:00, 13 Jun 2004 (UTC)
Readers will read the article, then decide if they're interested enough to look outside WP for more. Presumably is useful for more than the single statement; if not, it might be better to refer to a more generally informative web site instead. If by its nature the statement only comes from the one place, then it's more like a quote rather than received knowledge, and for quotes I think the inline ref is OK.

First of all, please don't confuse the full citation listing (Janet Q. Webauthor, (1999) "My favorite animal pests", retrieved June 2, 2004 from, which belongs at the end of the article, with the in-text pointer (Webauthor, 1999). The former is necessary to give a full description of the reference, especially if the website disappears in the future. The latter is often not necessary, but is sometimes useful if it is not clear which reference the reader should look up to find more information on a specific topic in the article.

Second of all, please don't use the current [1]-style auto-generation of web link references, which was very poorly thought-out. First, it is inconsistent, because a reference is treated differently just because it happens to reside online. Second, the autogenerated reference, with just the URL and no descriptive information, not even a title, is insufficient. —Steven G. Johnson 14:39, Jun 12, 2004 (UTC)

&sup1-3 and above

Question: (¹ ² ³ &sup4; &sup5; &sup6;) Why does the footnote solution work through #3? Am I doing something incorrect? Trc | [msg] 00:27, 15 Jun 2004 (UTC)

References/Books/Further reading/Bibliography (from Village Pump)

There seems not to be any standard regarding the terminology of what I would prefer to call "Further reading" at the foot of articles. At the moment, there is a wide variety of headings in use.

The "standard" is currently References. Whether people follow this is another matter. (For a reader, there is essentially no distinction between sources for further reading and sources from which the information was derived, and we shouldn't impose an artificial one.) —Steven G. Johnson 01:37, Jun 22, 2004 (UTC)

Secondly, there are many (especially history/biography related articles) incorporating 1911 Britannica text which have also incorporated wholesale the 1911 bibiography. With the exception of primary sources (which should be separately noticed and headed), I think these should be removed.

And thirdly, many contributors use headings conflatable with "Further reading" etc, usually headed "References" to note the publications from which they derive the information. These, I think, ought to be the province of numbered footnotes.

Whatever may be decided, I do think that a degree of standardisation is advisable. Djnjwd 22:58, 21 Jun 2004 (UTC)

Djnjwd, I have been trying to get people interested in the Wikiprojects concept for several days now, so that a group of people working together on the same subject can 1) follow the same formatting, as you suggest - ie Bibliography instead of Further Readings where applicable, External Links vs Resources, etc etc... and also so that 2) Categories can be smoothed out, lists can be made complete, the linking between categories is more organized and rational, etc. So far I have gotten little attention and only noticed arguments over what kind of categorization system to use... so it might be a while yet until your suggestions are noticed. It seems very hard to bring people together here on a specific topic, and there are also members opposed to such organization. My only answer for now is to encourage people you know are working on the same subject, to follow a single format/ structure. -- Simonides 00:02, 22 Jun 2004 (UTC)
What I personally think is that there should be an automatic way to generate numbered footnote references. E.g.;
     X is Y <ref>Z - How X is Y (1996), ISBN 0123456789</ref>,
     P isn't Q <ref>W - How P isn't Q (1997), ISBN 9876543210</ref>
would display as:
X is Y [1], P isn't Q [2]
and at the end of the article there would be an auto-generated references section with a list:
  1. Z - How X is Y (1996), ISBN 0123456789
  2. W - How P isn't Q (1997), ISBN 9876543210
There's a billion ways formatting and syntax could be done, of course. Fredrik | talk 23:13, 21 Jun 2004 (UTC)
The whole auto-numbering question has been discussed ad nauseam on Wikipedia talk:Cite sources. One problem with your style of proposal is that you assume each reference will be cited in the text exactly once. A more typical case is for a reference to be listed at the end only. (Or a reference might be pointed to from the text more than once). In any case, the main problem is getting people to cite good references in the first place; the whole auto-formatting argument is a red herring. —Steven G. Johnson 01:24, Jun 22, 2004 (UTC)
Wikipedia:Cite sources seems to come down on the side of ==references== rather than any other header. Yes, an automatic way would be ideal. You'd certain get my vote. Dieter Simon 23:47, 21 Jun 2004 (UTC)
I am quite opposed to the idea of using ISBNs, because they are misleading. An ISBN usually points to a single edition of a book, not to a title - the edition may be out of print, it may only be available in a certain country, it may be more expensive, it may be harder to find than other widely available editions, it may be out of date, it may not be in the original language, and so on. I think everyone should include only the title and authors, and the year of printing should refer either to 1) year of copyright on latest paperback copy or 2) year of copyright on first printing (perhaps both) and the language should be specified (ex. do not include original copyright year of German publication for a translated English title.) -- Simonides 00:02, 22 Jun 2004 (UTC)
Citation in academic work (essays, theses, research work, etc) does indeed normally preclude ISBN numbers for precisely the reasons you have given, why should you in a work of reference, Simonides? So, I suggest don't use these numbers. As long as you give the details of the particular edition/publication of the work cited plus all the other details, all should be well. Dieter Simon 01:04, 23 Jun 2004 (UTC)
An ISBN may in some sense be misleading, but in the case of some book-related servers, interconnections are easily accessible. Titles and authors are hyperlinked, for example. Usage of ISBNs facilitates book finding quite handily in this environment. Trc | [msg] 05:25, 23 Jun 2004 (UTC)
Linked ISBNs fit nicely with the hyperlink concept -- they're the closest we can get as long as books are not published free of charge on the Internet.
If a book has been published in many editions, there is probably something at least slightly notable about it; in that case I suggest writing a separate article about it and putting the ISBNs there. Fredrik | talk 11:05, 23 Jun 2004 (UTC)

Citing legal sources

How do we do this? I can't find much info on Wikipedia about this and there are several articles that need proper referencing. - Ta bu shi da yu 12:45, 30 Jun 2004 (UTC)

How is it done in print? And online? Is there an issue with Wikipedia copying what is done elsewhere? Citing science papers is the same on WP as anywhere else. You could also ask User:Alex756 - he is a lawyer so may have an opinion. Pcb21| Pete 20:26, 30 Jun 2004 (UTC)
You might want to take a look at Brown v. Board for an example. -- Jmabel
In the citation of Supreme Court cases, it is sufficient to name them by their case titles alone. For example, Brown v. Board of Education is sufficient. There is no need to add case numbers, etc. For Congressional documents, it is sufficient to cite them by their formal names. For example, Patsy T. Mink Equal Opportunity in Education Act. It can also be referenced as their official legislative file numbers. For example, the Hawai'i Overthrow Apology Resolution passed by a joint session of Congress can be referenced as United States Public Law 103-150. Statutes of individual states and charter amendments of individual municipalities and counties should be referenced in the same way. Gerald Farinas 15:02, 1 Jul 2004 (UTC)
I've discovered the following PDF document for Australian legal citation: - Ta bu shi da yu 12:36, 4 Sep 2004 (UTC)

We should make a formal standard for this. I'm actually in favour of giving a full citation for cases which are the subject of the article, as is done at the start of Brown v. Board of Education. A gray area for me is whether or not to give cases which are mentioned in the body of the article a full citation. Most legal style guides (including the one from my University that Ta bu shi da yu linked) would put the case name in the body of the text and the full citation in a footnote. That would be ideal, giving full citations for each of the numerous cases mentioned in an article looks cumberson (see brown again as an example) but making lots of footnotes is not only time consuming for the writer but might look cumbersome too. Whatever is felt, there should be a proper recommended style and it should be on the citation page. Psychobabble 07:19, 28 Nov 2004 (UTC)

Legal citation is an encyclopedic topic as well as a MoS concern

I think we probably should start articles about legal citation for each country -- after all, it's an encyclopedic topic -- and then base our standards on our own research. We should be covering both statute and case law. Right now Court citation has good material on the U.S. (and to a lesser extent Canadian) case law, but nothing on citing statutes. I've started German legal citation because I needed it when translating Paragraph 175. Other thoughts? -- Jmabel 18:50, Sep 12, 2004 (UTC)


I'm a little confused by what capitalization we should follow for the reference links at the end of an article. The "Books" examples only have the first word capitalized, even though it says we should follow the APA style. The "Journal Articles" section says specifically not to capitalize the article title, but the example leaves the journal title capitalized.

Is this to conform with the Wikipedia article titling rules? The thing is that the Wikipedia:Naming conventions (capitalization) page says you should capitalize proper nouns, and leave book titles, so if these articles, journals, and books are in Wikipedia, they're probably going to be capitalized.

Should we just capitalize the titles of artices and books normally?

(Disclaimer: I added the "Examples" to the "Naming conventions (capitalization)" page myself, because I thought it would be helpful and they seemed to reflect common usage. The examples are thus not evidence on this issue ^_^)

Creidieki 01:03, 24 Jul 2004 (UTC)

Whoops, the first "books" example was erroneous; fixed, thanks. In general, you only capitalize the first word of article titles (including articles in a journal or edited book), but you do fully capitalize book titles (including the titles of journals, magazines, etcetera). This is not just Wikipedia article titling; it is standard for many citation styles, including APA. —Steven G. Johnson 01:52, Jul 24, 2004 (UTC)
Well, I've done more research. It appears that, in the APA style, capitalization is different between the body of text and the "Works Cited" list. In the "Works Cited" list, for everything except the name of a journal, you only capitalize the first words of titles, proper nouns, and things after a colon. In the body of the text, and for journal titles in the Works Cited list, you capitalize all "important" words. MLA and Chicago want all important words capitalized both in the text and in the references.
So, according to APA, the first "books" example was actually correct. According to the others, I think you want the article titles capitalized in both. Personally, I think the APA method is silly in this case, and that we should capitalize all of the important words of both titles. The name of the article and the name of the journal are both proper nouns.Creidieki 02:53, 24 Jul 2004 (UTC)
Pardon the AOL-ism, but I agree with you Creidieki. DocWatson42 11:07, 15 Aug 2004 (UTC)

Displaying URLs separately

I've changed the standard for citing websites, because it was inconsistent with guidelines stated elsewhere which take into account the print stylesheet which writes URLs out in full when the article is printed. gracefool 11:06, 24 Jul 2004 (UTC)

Sources in foreign languages

I'm using a pair of books I brought back from Peterhof to expand the article about that location. I figure the citations should reflect the fact that they're written in Russian. How do I do this? --Smack 23:46, 29 Jul 2004 (UTC)

If you give a citation for the book with the title in cyrillic characters (and probably with a publisher located in Russia), it should be obvious to the reader, right? --Michael K. Smith 15:22, 31 Jul 2004 (UTC)

tool for citing sources

redlightgreen is a free tool for citing sources and can also be used to gather ISBN numbers for each specific edition of a book (shows all editions). it can reference in MLA, APA, Chicago and Turabian styles.

I think that Wikipedia should look to it's own way of citing based on one of the existing standards. when citing I also give an isbn number, providing a quick way to buy the book if someone is using Wikipedia for research purposes.

lastly if Wikipedia ever make it into print citations will be more usefull than hyperlinks. Ohka- 22:02, 30 Jul 2004 (UTC)


User:Stevenj, your revert was confusing. Are you against wikipedia having guidelines for citations? You quite obviously did not read the link you removed, as it does not simply "deconstruct" authorship, but discusses issues and guidelines regarding the quality, rather than simply the style, of citations. The readdition of Weasel word was unecessary as Avoid weasel terms is already linked to, in addition to creating a two word section (not even one sentence). Also, please see: Wikipedia:Village_pump#Cite_sources. Hyacinth 21:30, 3 Aug 2004 (UTC)

I'm not against guidelines for citations, but I don't think your one-sentence guideline was worth a separate section, nor did it seem to add much compared to the introduction which also contains guidelines.
Regarding the external link, I did look at it, and it is mainly deconstruction of the social implications, etcetera of citations...this is not a criticism, I just don't think it is very helpful here. Moreover, what guidelines are given are often not very relevant to an encyclopedia, where citations are much different from in a journal article.
Regarding the weasel terms, I already tried to delete it and Dieter Simon gave an argument for why both links should be there...see the edit history. —Steven G. Johnson 22:55, Aug 3, 2004 (UTC)

Proposed guidelines

If we are to have a informative, verifiable, and neutral encyclopedia, then, in my opinion, we need to think more seriously about citations:

I don't agree that primary sources should be preferred. This is true for original research on a topic, but in an encyclopedia that only summarizes the existing views of a subject, a reference to a more-accessible review/reference work, such as a textbook, is far more appropriate. We are not trying to contribute new scholarly knowledge here, nor to participate in debates...only to summarize. —Steven G. Johnson 22:47, Aug 3, 2004 (UTC)
Wikipedia is a secondary source, but that doesn't mean we can't, or shouldn't use primary sources. Hyacinth 22:58, 3 Aug 2004 (UTC)
Primary doesn't mean better. Primary sources are good if you are trying to contribute new commentary or understanding on the history of a subject, but that's not the purpose of Wikipedia so this advantage is not so relevant. In fact, primary sources are often harder to find, harder to read, and harder to connect with existing knowledge than a good textbook or review...which is why the latter are often preferred. —Steven G. Johnson 23:10, Aug 3, 2004 (UTC)
Yes, I see now, thanks. How about now? Hyacinth 23:16, 3 Aug 2004 (UTC)
I don't understand this explanation, and I don't think that there is any reason to discourage citation of primary source material on Wikipedia. Wikipedia:No original research, a principle with which I think we all agree, requires Wikipedia articles not to be primary sources, but it doesn't and shouldn't require articles not to cite them. Also, Wikipedia can (and should!) present new commentary, summary, and synthesis of previously existing knowledge, legitimately and without conflict with Wikipedia:No original research. Don't confuse citation of primary sources with creation of primary research. -- Rbellin 05:10, 4 Aug 2004 (UTC)
I'm not confusing creation and citation of primary sources. Read what I wrote: I explicitly pointed out several of the disadvantages of citing primary sources in an encyclopedia. In contrast, the advantages of primary sources are often relevant mainly to original research. —Steven G. Johnson 15:50, Aug 4, 2004 (UTC)
Having read your comments, I disagree and feel that Wikipedia should, in many cases, encourage the citation of primary sources. I do not feel that Wikipedia's goal to be a reference work requires it to prefer to cite reference works. The reasons you give are good reminders that Wikipedia should aim for a general readership, but the goals of a "Further reading" section for general readers' information and a "References" section for verifiability and citation may, in some cases, conflict. And in such cases I'm in favor of retaining both, separate, priorities for the article. The terms "primary" and "secondary" are general and vague enough that we may be operating with different understandings of them. An example of why I feel primary source citations are important: Think of articles about philosophers and thinkers. In an article on the work of a philosopher, I think it borders on the ridiculous to prefer a citation to a biography of the thinker over a citation to his work. -- Rbellin 16:22, 4 Aug 2004 (UTC)

I would suggest something much shorter, so that it can go directly in the introduction, such as:

It is often preferable to cite recent authoritative textbooks and/or review articles instead of large numbers of primary sources, as the former are often more accessible and will usually have exhaustive citations themselves. Remember that Wikipedia is not for original research, so primary sources are not necessarily better.

—Steven G. Johnson 23:26, Aug 3, 2004 (UTC)

Proposal A Proposal B
It is often preferable to cite recent authoritative textbooks and/or review articles instead of large numbers of primary sources, as the former are often more accessible and will usually have exhaustive citations themselves. Remember that Wikipedia is not for original research, so primary sources are not necessarily better. Citations should be chosen and evaluated with respect to appropriate authority, timeliness, relevance, and Wikipedia:Verifiability."
Secondary sources are preferred to tertiary sources.
Contemporary primary sources are preferred over later primary sources.
The most recent secondary and tertiary sources are preferred over older secondary and tertiary sources.
See also: Wikipedia:No original research, Wikipedia:NPOV_tutorial#Attribution_and_citation, and Wikipedia:Check your facts.
Word count: 48 Word count: 56
Total: 104

Not only is B longer, but it is also more complicated and less colloquial. I think it's counterproductive to be that detailed and formal, because it will scare people off. People shouldn't have to worry about whether a source is secondary or tertiary, or which things are contemporary to what, they should just use their judgement about whether it is authoritative and helpful. (To me, the main need for a guideline is just to discourage people from thinking they should cite reams of primary sources.) —Steven G. Johnson 23:48, Aug 3, 2004 (UTC)

A is better. It wouldn't hurt to add a warning that primary sources often assume context and can be misinterpreted by the unaware; secondary sources are useful in supplying the context. Stan 04:05, 4 Aug 2004 (UTC)

A is not guidelines, however, but a guideline. Hyacinth 20:47, 4 Aug 2004 (UTC)

External link

  • The Citation Functions: Literary Production and Reception by The (In)Citers1, featuring full position statements and citation bibliography (1. These statements will be part of a roundtable discussion at the University of Tulsa's 1998 conference, "The Sociomaterial Turn: Excavating Modernism," held March 5-7.)

Hyacinth 21:30, 3 Aug 2004 (UTC)


How long will the style guidelines already in the article remain "proposed"? Will they ever be "adopted", and if so, how? Hyacinth 21:30, 3 Aug 2004 (UTC)

I think people will have to vote with their feet...i.e., when a citation style becomes anything like consistent in Wikipedia, it can be formalized. Moreover, calling it merely "proposed" serves an important purpose — it prevents people from worrying about formatting details too much, when right now (and for the immediate future) the quality of citations (or just having them in the first place) is far more important than formatting. It's not worth worrying about voting, or whatever, on an "official" style at this point. —Steven G. Johnson 23:07, Aug 3, 2004 (UTC)

Templates for citations

Readers of this page might be interested in some new templates that I have just opened up. Their aim is to make citation style easier to standardize. See Template:Book reference and its talk page. There is an example usage at Template:RefAudubonMarineMammals. Pcb21| Pete 09:19, 17 Aug 2004 (UTC)

Ok to Use Rumors as Sources in Wikipedia?

On the main page the anniversary of the patenting of the spork is listed, along with an illustration. However, I was dismayed to find this in the article:

"According to a rumor, the spork was invented in the 1940s by the United States Army, which introduced them to occupied Japan. It was hoped that the use of the spork would wean the people there from the use of chopsticks. This pointless hope did not come true; yet the spork that was spurned by the Japanese found a home in the United States of America, where its versatility and disposability were well adapted to the cuisine of the United States."

Am tempted to delete it, but maybe not? Is it really okay to cite a rumor as a source? Seems like there ought to be some factual basis, not just "according to a rumor..." H2O 07:07, 11 Aug 2004 (UTC)

If its a notable enough rumor (for example the folk etymology of posh is quite famous and has been published in books and such), its probably encyclopedic as a rumor, it should be included and noted as a rumor (hopefully with some reasoning for how it started/spread). Of course if a rumor is fact, it should be given as fact. In this case, there should be some more investigation probably. siroχo 07:42, Aug 11, 2004 (UTC)
I looked around and everything seems to point back to one guy's comment on a spork newsgroup some years ago. Hardly encyclopedic. Probably another urban legend. I deleted the rumors. If someone wants to verify this with something more than some newsgroup chatter, fine. H2O 07:48, 11 Aug 2004 (UTC)
If you are feeling confident about your facts and if you think it is important, then you could debunk the rumor as a rumor on the page itself. "Many web sites indicate that the spork was unsuccessfully introduced in Japan following WWII. However this rumour appears to originate from a single newsgroup posting [here]." Pcb21| Pete 10:06, 11 Aug 2004 (UTC)
I am not confident enough in my "facts" to include a debunking of a "rumour" in an encyclopedia article. That would be like starting another rumour. However, I have enough doubt that I think the rumour should be left out of the article until more evidence is available. Maybe someone knows of a person who lived or served in Japan around that time or has more knowledge of WWII history than I do. They could confirm or deny this rumour. H2O 15:58, 11 Aug 2004 (UTC)
Proper treatment of rumors, which by nature accumulate small mutations, includes IDing any near-truth in them, e.g.:
  1. In "the anniversary of the patenting of the spork", almost certainly distinguishing that from "... of a new design for a spork". ("Prior art" aside, popular culture has a pathetic misunderstanding of the incremental nature of invention and patenting.)
  2. Thinking not in terms of whether sporks were invented for Japan, but of whether there was a specific plan to introduce them there with the intent already stated.
And don't forget to copy this discussion to Talk:Spork.
--Jerzy(t) 17:26, 2004 Aug 11 (UTC)
I think it is fine as long as you can cite a good source describing the rumor (i.e. it is not just a rumor started by a random Wikipedian). e.g. "One rumor, according to the American Dictionary of Slang (1983), is that the "spork" originated as...." —Steven G. Johnson 22:39, Aug 11, 2004 (UTC)
Because incorrect folk etymologies accumulate around works like spork, I think that it is much better to discuss it intelligently within the article rather than to leave it out. Include rumors from suspect sources if you don't have better evidence and if you can establish the rumor as widespread, and verifiable, and label it as such. E.g. do a Google Groups search for it. If you turn it up, say, "The story that the spork originated in thus-and-such way has been widely repeated on the Internet in the USENET newsgroups. However, the first such mention is in the year 1998, and the absence of mentions prior to that time makes it unlikely ..." Readers can judge for themselves whether they trust USENET or like your methodology. Later on, if someone finds a better piece of information they can replace yours. If you just leave it out, people will keep reinserting versions because everyone wants to know the origin. If you can find a dictionary that says "origin uncertain" be sure to say "The so-and-so dictionary says origin uncertain." Say what you believe about the rumor, give verifiable reasons for your belief, and supply information that lets the reader judge the soundness of your statement. My $0.02. [[User:Dpbsmith|Dpbsmith (talk)]] 13:32, 12 Aug 2004 (UTC)

Non-English references in the English Wikipedia (copy from Village Pump)

I received a complaint that I used Dutch references in an article in the English Wikipedia. I used it because that is was all that I had. I also regularly add German and sometimes French references to the English Wikipedia when I do not have access to an English version. I try to replace them as soon as I can. Using non-Dutch references is very common in Dutch scholarly tradition. Is it allowed here? Thanks in advance. Andries 17:03, 15 Aug 2004 (UTC)

If you have the choice the english sources are preferred here, as that'd be the one easiest accessible to most user here. However if there is no such source (e.g. if you write about a Dutch city and add the dutch website as your source) it is better than none, you just should add a short remark about the language - a "Dutch" in brackets is enough. Wikipedia:Cite sources does not mention anything about non-english sources, so there seems to be no official policy. andy 17:32, 15 Aug 2004 (UTC)
"Using non-Dutch references is very common in Dutch scholarly tradition." That's because there is hardly such a thing as a monolingual Netherlander; and I sometimes feel like there is hardly such a thing as a multilingual American or Brit. Clearly, given the number of monolingual English-speakers in our readership, English-language references are strongly to be preferred. On the other hand, I myself have probably cited about 500-1000 non-English web pages in the last year. When I place them the "external links" section of an article, I always note "in Catalan", "in Romanian", "in German" etc. I try to avoid citing them in passing in the body of the article, because explanatory language is hard to insert without breaking the flow, and a lot of people will be annoyed to click to something they can't read. If you have to do this, it's best to do something like (as discussed on IDESCAT's Catalan-language site []) ==> (as discussed on IDESCAT's Catalan-language site [2]) instead of just [] ==> [3] -- Jmabel 06:37, Aug 16, 2004 (UTC)
Agreely strongly with all that of that, but I just wanted to emphasise that it is standard scientific even in English practice to quote foreign language papers if that is the best reference on a particular point. Wikipedia is less specialist; but shouldn't be afraid to do the same where appropriate. Pcb21| Pete 07:59, 16 Aug 2004 (UTC)
Use the best reference, whatever language it's in. English scholarly papers frequently cite papers written in languages other than English. As a courtesy to English-speaking-only readers, a short phrase describing the article in English wouldn't be out of line, and, of course, including similar references in English would be fine. There is a legitimate objection to articles in the English Wikipedia that are written entirely in languages other than English, but a complaint about a reference sounds inappropriate to me. The article in question gives three references, two in English. Without knowing the complaint I can't judge, but I'd be inclined to shrug it off as xenophobia. [[User:Dpbsmith|Dpbsmith (talk)]] 12:24, 16 Aug 2004 (UTC)

Sources and articles should be separated in MediaWiki 02:28, 19 Sep 2004 (UTC): As anyone who spends any time with wikipedia knows, conventions within articles for credited sources range from exhaustive and intrusive to suspect or nonexistent. As anyone who uses google knows, wikipedia and its "licensees" are appearing more highly-ranked in response to searches. These two facts are eventually going to make wikipedia a victim of its own success: the greater prominence of wikipedia will cause more people to refer to its content, yet the lack of any requirement for adequate documentation of source material will render some articles suspect and through guilt by association, color people's perceptions of other articles.

One answer to this is editorial approval of content changes. You see versions of that in other colloborative efforts such as or, but there are good reasons why wikipedia won't follow either of their models. The community has undertaken a debate about Wikipedia:Approval mechanisms which may lead to a consensus about how some form of editorial control gets introduced.

Regardless of what happens in the approval mechanism debate, I think first class support for citations is required. MediaWiki needs to be enhanced in a way that tracks source material for articles. Each version of an article needs to have a set of citations that gets maintained in parallel with the article itself.

A simple form of this is to have the citation page be nothing more than editable text. But I'd like to argue for a more structured approach, which contributors would fill out when they save a new page or article. A form supporting each style of citation mentioned in this project article could be filled out by the user. But saving them in a structured form, you'll make it possible in the future for all sorts of neat tricks, such as automatic validation or comparison of contributions with sources (for web pages), automatic inclusion of ISBN numbers (for published works), etc.

The importance of sources should be demonstrated in the user interface as appropriate. For example, in the monobook skin you would get a fifth tab at the top of the page to go along with article, discussion, edit this page, and history.

What made me suggest this after having contributed to wikipedia anonymously since May 2003 (over about 350 online sessions), is the issue of reliability of sources from the world of genealogy. If you try to collect your family history, you don't bother at first recording sources. But as your list of ancestors and relatives grows, you start to get conflicts in some data, you begin to wonder who it was that told you about great-great-great uncle so-and-so who settled the land that the old farm was built on, or you find it hard to corroborate the oral histories you've recorded with stuff you find in cemeteries and Pretty soon you don't know what to trust any more.

References vs. External Links

I'm having trouble figuring out the difference between "References" and "External Links". I see a lot of articles using External Links exclusively, simply because all of their references are online. Is there a difference between the two categories? I couldn't find any pages that talked about this. -- Creidieki 22:35, 22 Sep 2004 (UTC)

In my opinion the major difference between anything and a reference or source is that a reference is cited in the article. Thus an "External link" may be a reference, if cited. A link to an informative article that is not cited is never a reference or source, but simply an "External link" or at best "Further reading". Hyacinth 20:54, 4 Oct 2004 (UTC)
That may be your opinion, but it doesn't really correspond to actual usage. In something like an encyclopedia that is summarizing well-established knowledge, in-text citations are often superfluous, but references are still important. (I've been involved in print encyclopedias that used "References" sections in precisely this way.) In the real world (i.e. most scholarly articles that have reference lists), there's usually no clear-cut distinction anyway between references that are "sources" and those that are "further reading" — most references serve both purposes. —Steven G. Johnson 05:48, Oct 5, 2004 (UTC)
I think that making an artificial distinction between online and offline references is a bad idea, and I would therefore discourage External Links sections in most articles...but it's not really an important enough issue to spend much effort fighting for consistency on. Realize that the various section titles are partly a historical artifact, and don't try to hard to retrofit a coherent explanation onto pointless distinctions. —Steven G. Johnson

Speaking of opinions: "Actual usage" is not always preferred usage. What you seem to propose, a complete lack of distinctions between references, sources, further reading, and citations, allows little to no verifiability. I am not arguing for an arbitrary distinction between on and offline "references" (ie, further reading), and thus am little interested in that debate, I was simply arguing for verifiability. If you reread my above comments you'll realize I don't care how "sections" are labelled. In response to the following comment:

"In the real world (i.e. most scholarly articles that have reference lists), there's usually no clear-cut distinction anyway between references that are "sources" and those that are "further reading" — most references serve both purposes."

I propose that you check your condescension at the keyboard. I live in the real world, if that doesn't go without saying, and I've read real books. More importantly, you have created a disagreement where there is none. Hyacinth 18:42, 5 Oct 2004 (UTC)

You may not care how the sections are labelled, but I disagree that the references should be broken into sections in the first place. You seem to think that sorting a bibliography into "sources" and "further reading" (plus "references" and "citations") is helpful in verifying an article. For one thing, as I said, the distinction is typically artificial, and a source that is "further reading" can often equally be used to verify an article, even if it's not a "source" that was originally used to write the article. For another thing, if it is really important to let the reader know which particular reference is the basis for a particular statement (e.g. if a particular fact is found in only one place, or if sources disagree), that is the circumstance where an in-text parenthetical citation (Foobar, 1973) is appropriate — having separate lists of sources and further reading is immaterial, because you still need the in-text citation to find which source goes with which statement. —Steven G. Johnson 20:09, Oct 5, 2004 (UTC)

To me, "external links" are "what I can click on right now" while "references" are "visit the library or bookstore". I suspect that if you officially merged the two types of entries, after a while editors would get tired of hunting for clickables among the book titles, and would spontaneously start to segregate the two kinds of things, eventually titling each group, and voila, the wheel is reinvented. Compare to "further reading" vs "references" - editors have no idea how to characterize one vs the other, so typically they all get thrown into a single section. Stan 19:16, 5 Oct 2004 (UTC)

I agree exactly with the sentiment in your first sentence, except that for me, substitute "Further reading" for "References". To me, "References" means "list of specific citations for specific facts cited in the article". If you look at a real scholarly book, in addition to references (usually called "Notes", or "Footnotes" - although academic papers invariably call them "References", go figure), it also has a section called "Bibliography", which is more akin to our "Further reading" sections. I would strongly oppose use of the term "References" for anything except i) lists of specific sources for specific statements, or ii) definitive reference works (e.g. the "PDP-10 Processor Manual" on the PDP-10 page).
(a) In the textbooks I have on my shelf (e.g. Jackson, "Classical Electrodynamics"), typically each chapter has a "cited references and further reading" section, and sometimes there is a bibliography chapter at the end of the book...but in this case the bibliography is just the full citation for abbreviated citations (e.g. "Landau and Lifshitz") in the aforementioned sections. Furthermore, in shorter scholarly publications (e.g. journal articles), references "used" in creating the article are almost invariably mixed inextricably with things included for context or further reading. So, I don't really agree with your first statement. I don't think we should cite any source that isn't credible and respected, but requiring it to be "definitive" is going too far — for example, listing an introductory textbook is likely to be very helpful to a beginning reader, but an introductory textbook is unlikely to be the "definitive" work in a field, nor need it be. —Steven G. Johnson 00:34, Oct 29, 2004 (UTC)
Your textbook example matches my wishes exactly - specific citations go in "References", and the example you cite (an introductory textbook) would go in the "Further reading" section. Noel 01:48, 29 Oct 2004 (UTC)
I agree that an unambiguous breakdown into "references" and "external links" is possible according to whichever happen to be available online. However, I haven't had any practical problem in scanning a list of references and spotting the, I'm not sure that the problem that this solves (spotting the links) is so great. Furthermore, if there is an in-text citation to a source, it makes things harder if you have to go through to lists to find Foobar (1973) than if you have to go through one. Furthermore, the implications of saying that a source is not a "reference" just because it is online irk seems to imply a qualitative distinction in the content and not just the location, which isn't necessarily the case. And should we continually re-sort the two sections as references become available (or unavailable) online? —Steven G. Johnson 20:09, Oct 5, 2004 (UTC)
As in all things, there are exceptions. If an article like Linux has a very long list of web sites that are loosely associated with it, but aren't directly related to the article content per se, then a separate external links section may make sense. But as a general rule I don't think separate sections based on the location of a citation are worthwhile. —Steven G. Johnson
(See also my commments to Stan Shebs, above.) I would strongly oppose melding "Further reading" and "Links" not just for the immediacy, but because "Further reading" implies a level of quality control (as is usual with printed books, etc), permanence, etc that is one level above many (most?) web pages. How many times have you clicked on a link and found that the target wasn't there any more? Now, how many times have you gone to find a book listed in a biblio and found that it doesn't exist in the world anymore? And I won't even get into the editing, etc, etc. Look, I have nothing against the Web (see my bio :-), but in practical terms there is a real difference. Noel 18:48, 28 Oct 2004 (UTC)

Online references - valid?

Some time back, when I nominated a bunch of my works on WP:FAC, they received a great deal of support. However, there was one objection which was never removed (this held up I Want To Hold Your Hand; even when I personally contacted the objector on his talk page after fixing the objection, it was not withdrawn, nor was any reason given). The objection? That I made too liberal use of online references (i.e. Hey Jude and Something, both already featured articles). I think this should be resolved in policy, somehow. I mean, if it's some site on Tripod with no bibliography or references, and practically unknown anyhow, that's okay, it's reasonable. But the sites where I got my material from do list their references, and both score extremely high on Google when searching for "the Beatles". This site is the second result for "the Beatles", just behind the official Beatles website. This site provides a bibliography. Clearly if the references are proven credible, they should be allowed to stand, whereas referencing some teenybopper's Frontpage-designed Geocities site shouldn't. Sorry if I'm rambling here, I'd just like to hear others' opinions on this. Johnleemk | Talk 15:39, 18 Oct 2004 (UTC)

I don't know about FAC, but I don't see why online references should inherently be a problem. Maurreen 02:20, 29 Oct 2004 (UTC)
(William M. Connolley 11:19, 29 Oct 2004 (UTC)) Me neither. I use them extensively on, say Global warming. Online refs (if respectable) are more useful to 99% of people than offline ones.
I don't think there is any general problem with online references. On the other hand, it is desirable to check an article against professional, authoritative, reviewed, and time-tested sources, and in many (not all) cases these reside offline. When all of the citations are online, there is sometimes a knee-jerk reaction (perhaps unfair, in your case) that the sources are superficial or non-authoritative. There is an argument that looking at a few offline sources is a good idea, because (a) it makes sure you haven't missed some systemic weakness in the online references and (b) it combats the perception that the article is weakly researched (even if it's not the reality in your case). —Steven G. Johnson 21:44, Oct 29, 2004 (UTC)

(Let me also sound a general note of caution about using Google rankings as a measure of a site's authoritativeness or accuracy. —Steven G. Johnson)


It's not a big deal, but I'd like to say why I had added this: "Sources are especially important when stating any opinion and stating anything that is disputed." There is a dispute concerning the Style guide article. I have asked for a source for more than 10 days. Maurreen 17:47, 31 Oct 2004 (UTC)

I agree that sources should be cited in disputes. The reason it's not important to admonish this prominently here is that almost everyone agrees on this. When you challenge someone on a fact, especially if you cite an explicit source that supports your position, they pretty much have to cite something to the contrary. The main problem comes from the fact that people often don't cite sources until they are challenged.
(In the Style guide article you mention, I notice that the other side has cited various places in style guides, and various quotations about them. You don't find these arguments persuasive, but that is another matter entirely. We can't simply say "make good arguments" here.) —Steven G. Johnson 19:29, Oct 31, 2004 (UTC)
Thank you for reinforcing the part about opinions. Maurreen 20:29, 31 Oct 2004 (UTC)

Journal citations

I've just been taken to task for doing these 'wrong', when I use the standard scientific journal style. Being referred to here, I found this style:

  • Brandybuck, M. (1955). Herb-lore of the Shire. J. Royal Institute of Chemistry 10 (2), 234–351.

What I am used to seeing in scientific journals (numerous examples) is as follows; I'd suggest we change to this style, which is the de facto scientific journal standard:

  • Brandybuck, M. (1955). Herb-lore of the Shire. J. Royal Institute of Chemistry 10: 234-351.

MPF 10:01, 10 Nov 2004 (UTC)

Actually, the standard format in all of the journals I use is something closer to:
  • M. Brandybuck, "Herb-lore of the Shire," J. Royal Institute of Chemistry 10 (2), 234–351 (1955).
Overall, it depends a lot on what field you are in. I don't really care too much what format we recommend, but I don't think we should keep switching back and forth as each person comes along and changes it to what they are used to in their field. (It's also important to give the option of including the issue number as well as the volume; some journals, for example the IEEE journals, are harder to look up online without the issue number.) —Steven G. Johnson 15:28, Nov 10, 2004 (UTC)
If it depends on which field you are in, I don't think we should force everyone to adopt the standards of a different field; botanists and zoologists (whose journals mainly use colons) shouldn't be required to use the standards set by arts & humanities (as far as I know where commas are more used). Point taken on the part number though, I've no objection to that being included. - MPF 16:18, 10 Nov 2004 (UTC)
We're not forcing anyone to do anything. (As the article stresses, the first thing we care about is that the references get cited in the first place; the format is a very distant second.) The suggested style is mainly there for people who aren't sure what style to use and want some help in picking one, and also to give an example of what information a citation should provide. The secondary purpose is to provide a starting point if people ever want to standardize the format in Wikipedia (a distant prospect, at this point). It's not helpful, however, to suggest multiple styles. —Steven G. Johnson 18:51, Nov 10, 2004 (UTC)
If it isn't compulsory, then it should be made much clearer that other widely used styles are also acceptable; the style sheet looks pretty prescriptive to me - MPF 19:11, 10 Nov 2004 (UTC)
I agree. If the citation format is a proposal, then alternative proposals should be described in the same place (rather than just this talk page); if it's a guideline, then it should be backed by some measure of consensus (more than I think there is, at least for the formatting details). However, it's also important to note that a guiding principle of this place is Wikipedia:Ignore all rules, and all "rules" should be considered in that light. In fact, the citation formatting guideline might profit from a link to that page. -- Rbellin 22:17, 10 Nov 2004 (UTC)
I'm going to third this. If some fields use ":", and that's acceptable here, then we should say so. The way to handle people who come to this page who don't know what to use, and are looking for guidance, is to say "if you don't know what to do, use this format", not leave all the other ones out. Now, perhaps we need to have a conversation before deciding which one to recommend. I have no opion there, personally. Noel (talk) 12:02, 19 Nov 2004 (UTC)
I'm fine with saying, There is currently no firm policy on citation format in Wikipedia, but if you can't decide what style to use or don't know what information to include, use this one. However, I am against giving examples in all possible citation styles in the article. What purpose does that serve? (a) This page is not a survey of styles (which might make a good Wikipedia entry, however). (b) The Talk page is for discussion and debate, while the article is to help contributors now. (c) Users who are used to an alternate style (like colons) already know the style, so this does not help them. (d) Putting long lists of alternate styles in the article just makes it longer and more confusing and therefore less helpful to users who don't know a style. —Steven G. Johnson 17:40, Nov 19, 2004 (UTC)
(And putting no style suggestion in the article until there is a Wikipedia-wide consensus is also not helpful to users. —Steven G. Johnson)

I don't think a Wikipedia-wide consensus is needed, just a consensus here. Maurreen 05:50, 20 Nov 2004 (UTC)

Is this policy or "Style and How-to series"?

Normally, articles are either "policy and guidelines" or in the "Style and How-to series". This seems to be in both. Please fix this, for the sake of the organization of the Wikipedia Namespace. Thanks! JesseW 14:01, 17 Nov 2004 (UTC)

This page is non-binding guidance. At least at present. This is evidenced by it being the in "Style and How-to" series, but not being in Category:Wikipedia official policy. The project page itself does not describe itself as policy. Also, if it were policy, it would be called Wikipedia:Manual of Style (citing sources) or the like. jguk 19:47, 17 Nov 2004 (UTC)
I've removed the line at the top that said it was part of Wikipedia policy and guidelines, so now it's firmly in the catageory of "Style and How-to series". So is the Manual of Style policy? I hadn't realized. What does that mean, exactly? JesseW 13:10, 18 Nov 2004 (UTC)

Wikipedia should allow MLA documentation

There's a site called EasyBib that makes it really easy to generate MLA-compliant bibliographies. I think that simply based on the existence of that site, Wikipedia should allow MLA documentation. I'm going to recommend the use of that site to newbies who aren't inclinded to write references lists all by themselves.

Vespristiano 18:48, 2004 Jan 10 (UTC)

It looks like this is a for-pay site?? I don't think Wikipedia should rely on such a thing. Besides, it's not like formatting references is hard. The main thing is to encourage people to enter the information; it can be formatted in a canonical way later, if necessary. —Steven G. Johnson 20:47, 11 Jan 2004 (UTC)

Why APA?

Why does the "proposal" for Wikipedia citation call for APA style? Without wishing this discussion to become pedantic, I'd like to observe that author-date citations are most useful for citing recent research by practitioners in a single scholarly field (in this case, the date is unambiguous and significant information). When citing sources that have been published in multiple editions and translations (historical documents, scholarly texts pre-1950, etc.), the date used in the citation (date of publication of the cited edition) can be very misleading to readers. In this case you often see things like (Hegel 1983) referring to recent translations of 19th-century texts. That's not appropriate for a mass readership, I'd say, and I'd propose that MLA style is easier on the reader in such cases. -- Rbellin 19:55, 25 Feb 2004 (UTC)

I entirely agree regarding the undesirability of adopting APA as the citation standard for a general encyclopedia. To quote from the back cover of the 4th ed., "The Publication Manual of the American Psychological Assocation is the style manual most used by writers and students in psychology, the other behavioral and social sciences, nursing, criminology, and personnel areas." And that's it. It was not intended for use in, and (in my opinion) has no place in, history, literature, religion, law, or the arts. I am also not aware of any subdivision of the physical or natural sciences (including medicine), nor of technology, where APA or any similar "short citation" system is used. APA was intended for use by a relatively small scholarly group for very narrow purposes. I object particularly to the notion of including only an author's first initial rather than the entire first name -- and especially for older sources. (Why would you want to cite a work by "Freud, S." or "Asimov, I."?) Having said all that, I personally would recommend the Chicago Style Manual (or Turabian, which is its subset) and/or MLA, both of which are intended for use in much wider areas of knowledge than APA. Michael K. Smith 16:44, 3 Jul 2004 (UTC)

Regarding author/date citations, the main common alternative is numbered citations, and the latter is impractical without technical support in the wiki software. Given author/date citations, it makes most sense to use a style organized with the author last name first, and date soon thereafter. Regarding using initials, note that the article already recommends using the full author names, or the names as printed in the source, since we are not limited in space. The recommendation was only for a style "based" on APA, not slavish dedication to APA guidelines. Anyway, this whole argument is besides the point, because hardly anyone follows any consistent citation style at the moment; the real battle is to get people to cite sources in the first place. —Steven G. Johnson 18:08, Jul 5, 2004 (UTC)

Of course, we all agree (I hope), it's most important to get citations into Wikipedia in any form whatever. But that doesn't prevent us from discussing what form they should ideally take, and eventually working out a good solution. Author/date (i.e. APA-esque) and numbered endnotes are not the only (or the only "common") alternatives. The two comments above suggest two or three other choices (Chicago, Turabian, MLA) which I believe to be better than author/date citation for a general readership and a broad range of topics. There are good reasons not to use author/date citation outside of journal articles in a single, relatively coherent scholarly field; the biggest, to my mind, is that it gets mighty confusing to all readers (even specialists!) when the date of the cited edition/translation doesn't match even the century of the original publication. -- Rbellin 05:20, 4 Aug 2004 (UTC)

I see your point about APA-style not being the best option for a general readership. I agree that this is a good reason for switching to an alternative style that's easier to understand. On the other hand, I think that your objection about dates of citations in text not matching the date of original publication is based on a misunderstanding. According to the APA-style manual (fifth edition, p.254-255) authors should also cite the original year of publication of a republished work. Example: Charles Darwin (1859/1998) wrote that sexual selection is an important selection mechanism. So, the issue with the years is actually caused by people *not* adhering to APA style. -- Sietse 11:43, 29 Oct 2004 (UTC)

MLA style is superior, IMO. [[User:Neutrality|Neutrality (talk)]] 20:37, Sep 20, 2004 (UTC)

I also think the APA style is ugly. For "Further reading" entries, I prefer some variant of the Turabian "Note Entry Form"; listing things as being by "Luser, J. Random" is so academic and ugly it makes me want to puke. Since when do we order "Further reading" entries alphabetically, anyway? I always put the most important ones at the top. Noel 18:18, 28 Oct 2004 (UTC)
To be completely honest, even though I proposed it, I find the APA bibliographic style a bit ugly too, along with MLA and Turabian too (they're all pretty similar in the bib. format). Coming from the physical sciences, I prefer something like the IEEE style, which doesn't put the last name first. However, I couldn't justify IEEE because it is designed for a numbered citation list, which we don't have technical support for; whenever you have (author, date) citations, listing by author and date seems to be the established practice. On the other hand, you could argue that an encyclopedia article should generally only have a handful of cited references, if that, so sorting the list in a corresponding order is actually not that important. Ultimately, I'm agnostic about the bibliographic format (I'd prefer BibTeX!). —Steven G. Johnson 15:25, Oct 29, 2004 (UTC)
Right, which is why I specified the Turabian "Note Entry Form", not their main citation variant, which also uses "Luser, J. Random." Noel (talk) 19:10, 31 Oct 2004 (UTC)
HOWEVER, the original poster's objection was not about the style for the bibliographic entries, but rather about the in-text citation style. There I have to disagree — I think (author, year) is most appropriate, absent technical support for a numbered reference style. MLA's style is (author, page). My objection to this is that the page in the text is just a random number that is irrelevant while reading the article, whereas the year provides some useful information about the currency of a given citation. (As Sietse pointed out, worrying about republication dates stems from a misunderstanding of APA. In any case, the in-text format is just a convenient abbreviation for the full reference at the end.) Moreover, a specific page number for a specific fact is unnecessary for most references anyway, unless a particular item is very hard to find; this facet of MLA is more suited to esoteric humanities publications arguing about very specific statements of various authors. —Steven G. Johnson 15:25, Oct 29, 2004 (UTC)
Right, I agree with you for in-text citations. In academic mode, I always hated the numbered style, because I always had to interrupt my train of comprehension to flip to the back to see if it was a reference I knew. With the name/date form, you just say to yourself "yah, I know that one" and keep going. However... I wonder if there's some way it could be made a user-selected presentation format, the way dates are now? I mean, the data would have to be stored (in internal format) in some way that linked the citation body to the location in the text anyway (so that when you add a cite between #7 and #8, it auto-renumbers all the ones through #77), so it should be possible. That way everyone could be happy! Noel (talk) 19:10, 31 Oct 2004 (UTC)
I disagree about the irrelevance of page numbers in citations, and believe they should be strongly encouraged. There are two reasons why: verifiability and ease of reference. Both a Wikipedian interested in verifying the correctness of a citation and a reader interested in finding out more about the subject are likely to be very interested in the specific page cited (and will often be unable to locate it otherwise). Remember, many Wikipedia articles cite books, not just short texts. -- Rbellin 16:12, 29 Oct 2004 (UTC)
Even in the case you describe, the page number is unhelpful in the text — it can still be in the complete reference at the end, if it is necessary in a particular case.
Nor did I say that page numbers were never useful, merely that that are not necessary most of the time. Which is true — because encyclopedias summarize only well-established knowledge, most articles will not be like a scholarly paper that picks one fact from here and and other from there ad nauseam. Rather, when you reference a textbook, for example, it will normally be for a textbook that has a general overview of the subject in question, not for a specific fact; the book's table of contents and index should usually be more than sufficient to let the reader find any particular bit of information they care to know.
Having the page number in the in-text reference is useful only when you need to reference multiple specific facts from a single source, and even in this case APA provides a way (author, date, page) to do it. But this need should be rare, for the reasons above. In fact, as discussed elsewhere in this talk page, in-text references should be rare in general in an encyclopedia article (as opposed to original research). —Steven G. Johnson 21:32, Oct 29, 2004 (UTC)
That last point -- on the desirable frequency of in-text citations -- seems quite contestable to me, but that should be a separate discussion and doesn't need to be broached now. I think there's consensus here on everything but the finest details. I agree with you that page numbers are not necessary in in-text citations when broad facts or general discussions are being drawn in summary form from a source. In this case, for readability, only the minimal citation information necessary to find a reference in the end-of-article list should be given (and, as an aside, I personally prefer MLA here, too, which allows you even more brevity: you can drop the date and just use the author name [using titles or partial titles to disambiguate if more than one work by the same author is cited]. In this simplest case, the date, too, is extraneous in the middle of the article text).
It seems to me we can begin to formulate a general agreement on some principles for in-text citation formatting, so I'll propose this as a possible addition to the guidelines below (and others can alter it, discard it in disgust, add it to the guidelines, or provide feedback as desired). At this point it seems less important whether we call this "MLA" or "APA" or "Turabian" or "Wikipedia" style than that we state some explicit guidelines for how, and when, and why, to cite sources in the text of Wikipedia articles. On the minutiae of formatting, we can split the difference for now and allow individual Wikipedians to proceed as they prefer. On the frequency of citations, since there is disagreement, there's no need for the guidelines to take any position on "in-text citations most of the time" vs. "some of the time" vs. "rarely" right now.
I would also like it if someone could emphasize the date (original publication vs. republication) issue (discussed above) in the reference-list formatting guidelines. I don't have the APA guidelines handy to read from (and again, Wikipedia can get it right even if APA is confusing on this), and I think this misunderstanding is all too common even among academics. Let's be clearer about it here. -- Rbellin 20:53, 6 Nov 2004 (UTC)

Proposed in-text citation guidelines

  • In-text citations should provide the minimal information necessary to find the fact or opinion being cited. This means:
    • If a summary of a substantial portion of a source, or a general argument, or a discussion of a broad subject, or a well-known fact is being cited, cite the entire source work in the article text. Use the author's name alone if there is only one work by that author in the reference list; otherwise, use the date of publication or the title of the work (in abbreviated form if desired) as well. So if Darwin's Origin of Species is his only work listed among the references, use "(Darwin)" in the text; if there are others, use "(Darwin 1859)", or "(Darwin Origin)".
    • If a passage is quoted directly, multiple citations to the same source are given, or it will be difficult for a reader to find the fact or discussion being cited, provide a page number as well as the minimal information above. Use "(Darwin 11)", "(Darwin 1859, 11)", or "(Darwin Origin 11)" according to the above guideline.
-- Rbellin 20:53, 6 Nov 2004 (UTC)
On the one hand, I like the idea of providing only minimal information. On the other hand, I expect that such a guideline will make in-text citations in large articles prone to ambiguities, especially in a collaborative project like Wikipedia. Example: User:Aaron creates an article about Charles Darwin and cites Voyage of the beagle as "(Darwin)". A few days later, User:Bert joins the fun and adds a paragraph about evolution theory, in which he cites The origin of species as "(Darwin, 1859)". Readers will first encounter the "(Darwin)" in-text citations; when they look up the citations in the reference list, they will have to deduce which of the two books they are being referred to. In my opinion, omitting years of publication should not be recommended, in order to avoid such problems.
I also propose that we recommend "p." as a prefix for page numbers so that it is clear what these numbers mean. This will be helpful for people who are not/less familiar with scientific citation formats.
Last idea: mention the republication date in the in-text citations too, especially if page numbers are provided. I think it is a bit confusing when page numbers refer to a republished edition, whereas the in-text citations suggest that they refer to the original work. Sietse 10:35, 22 Nov 2004 (UTC)

I too would like to see articles better cited than normal encyclopaedias (story of recent aggravation with one such elided), but I too don't like what in-text cites do to the flow of the text. However: Wikipedia is not paper; we can have our cake and eat it too! In the long run, I would like to see cites supported as first-class metadata, so one could e.g. click on a tab (as suggested by others) and see the articles with (copious :-) cites shown. However, in the interim, we have the source - we can put citations in as comments, something like <!-- Cite: Darwin, 1859; 1978 reprint pp. 172-173 -->. And if we fix on a set syntax for the leading tag for cites in comments, then if and when the glorious day finally comes that they are supported as first-class metadata, they can all be converted automagically by a bot. This is something we can start today - so can we agree on a syntax, and get to it? Noel (talk) 13:45, 22 Nov 2004 (UTC)

How 'bout "Cite:" as you suggested. I'll start using that, and mention it on my user page. If another one is picked, we can bot-change the "Cite" ones to fit it. JesseW 14:20, 22 Nov 2004 (UTC)
I think this is good as a general idea, but it seem like "Source" or "Reference" would be more clear than "Cite".
Also, I'm just curious about thoughts on footnotes. I'm ambivalent about them. Maurreen 16:20, 22 Nov 2004 (UTC)
It's just a hidden tag for in-text citations (e.g. "Brandybuck 77, pg. 129"), not anything readers would see. I suggested "Cite" because it's short and quick to type. Noel (talk) 23:17, 3 Dec 2004 (UTC)
The idea of citations (and references?) as meta-data sounds great to me. Maybe we could use a BibTeX-like syntax for this, so we don't have to re-invent the wheel? I think it would be best if we start adding such tags only when the meta-data system is working, otherwise articles would have either no visible citations (i.e. the citation is a comment) or double citations (i.e. both comment and text citations, so redundant) in the mean time. Sietse 08:56, 24 Nov 2004 (UTC)
I like the idea of BibTex for the list of references, because it explicitly tags volume/number/year, etc - which could (down the road) fix our problems with MLA/APA/Turabian/etc - eventually your reference list style could become a user option, like dates are now.
However, I think BibText only does the list of references at the end, not the individual citations in the body of the text, right? So we still need something else for those.
What's the deal on the meta-data system, anyway? Any idea when it's likely to arrive? If it's going to be a while, I'd still like to go with my original suggestion: pick a syntax for doing in-text citations as HTML comments now; they can be automagically converted (by a bot) to whatever the meta-data syntax turns out to be later.
I didn't understand your comment about "double or none" - I was assuming that for in-text citations there'd only be one, as a comment in the source (i.e. not visible normally). Noel (talk) 23:17, 3 Dec 2004 (UTC)

I don't like the idea of hiding the in-text citation in a comment. These pointers are there because they provide useful information — they tell the reader that a particular statement came from somewhere that can be easily looked up, they give some information about currency (via the year), and they tell the reader where to find the reference. (Yes, you can say that all of this will be available once again at some unspecified point in the future via improvements in the software. That doesn't excuse degrading the utility of Wikipedia now.) If you have so many in-text citations in your article that it is getting in the way of reading, then something may well be wrong with the article (e.g. it may be verging on original research). —Steven G. Johnson 00:27, Dec 4, 2004 (UTC)

Encyclopaedias generally don't cite in the text at all, so far as I know. All the rationales you cited above would of course equally apply to them, so they must have some good reason for not doing so. In addition, I think adding citations in the text that ordinary readers see would be obtrusive and distracting to them. Wikipedia is not a scholarly journal article, and our readers will not be (in large part) academics. Putting citations in the source, so they are available to editors, and for a meta-data system later, seems like a good balance.
Also, I am dubious about the "so many .. citations" point. I am involved in a number of pages on various topics where I see editors asking each other "what's your authority for that point" a lot. Many such articles are assembled from a long list of secondary sources, so it's hardly original research, but those sources may disagree. E.g. I'm doing some updates to the articles about British crypto efforts against the higher-level German ciphers in WWII (e.g. Tunny and Sturgeon), but my secondary sources often disagree, and because of that I'm always listing where I found stuff. Another example is in Egyptology, where many points are still not settled; e.g. how long Akhenaten and his father were co-regents. Even if you NPOV simply give all sides of the debate, you ought to provide a citation to each one. Neither one of these cases even comes close to the line of "original research": scholars in the field simply disagree.
Even worse than these easy cases (or so they seem, compared to what's next), there are many contentious topics (which are things I devoutly stay away from), of which the Wikipedia has plenty, where things will be even worse than this - people will want citations for everything. To not put them in because it would create an article that is too logged with citations is, IMO, a grievous error - going back and adding the citations later is an incredible amount of work, and frankly is just unlikely to happen.
Which gets us back to my previous point - hide them where the ordinary readers don't see them. Having them hidden in the source (available for later bot processing) is far better than not having them at all. Noel (talk) 00:23, 21 Dec 2004 (UTC)
"Encyclopaedias generally don't cite in the text at all." But encyclopedias generally don't have our complex issues of open collective authorship, and consequent need to validate that statements weren't pulled out of someone's ass. -- Jmabel | Talk 00:33, Dec 21, 2004 (UTC)
Sure, but that doesn't mean they don't contain errors. 47 Ronin currently contains a bunch of errors because I relied (in part) on the Japan Encyclopaedia, published by Harvard University Press. Citations would have been really useful there.
Anyway, the issue is not whether or not to add citations in the body of the article - I think we all agree that they are a good idea. The only question is "do they display for all readers, or are they hidden".
If my only choices are "only a few citations" or "hide them", the second is the only viable way to go, as far as I'm concerned. Noel (talk) 03:01, 21 Dec 2004 (UTC)

Adding bibliography guidelines: Any objections or comments?

I would like to add more information about APA style rules for formatting references and citations to the An example citation style section. Specifically, I want to

  1. add rules for citing republished or translated works, as well as 'secondary' works such as book reviews.
  2. add rules for references to non-standard sources like television series episodes
  3. explain various minor rules concerning things like capitalization and how to describe the location of the publisher of a book.

My additions will be based on the APA Publication Manual, fifth edition. Since this page is based on consensus, I would like to know in advance if anyone thinks that it is necessary to discuss this first (e.g. because it makes the page more complex). Suggestions, comments, etc. are also welcome. If I receive no replies within a week or so, I'll assume that it is okay with everyone and start changing it. Sietse 09:21, 24 Nov 2004 (UTC)

It sounds fine to me. JesseW 14:32, 24 Nov 2004 (UTC)
Fine with me, as long as you try to keep simple things simple. Be sure to come up with good jokes for fake citations. =) —Steven G. Johnson 19:49, Nov 24, 2004 (UTC)
Could you show us a draft here first? Maurreen 04:39, 25 Nov 2004 (UTC)

Related page

You might be interested in Wikipedia:Forum for Encyclopedic Standards. The group is discussing reference standards, possible software changes and the like. Perhaps there should be some coordination between here and there. Maurreen 15:47, 28 Nov 2004 (UTC)

Quotes around article names?

Shouldn't there be quotes around the names of articles (still with italics around the title of the whole publication)? --L33tminion | (talk) 22:18, Dec 3, 2004 (UTC)

In many style guides, yes. Not in APA style. (See e.g. [4].) —Steven G. Johnson 22:50, Dec 3, 2004 (UTC)

If the author cannot be determined...

The following was recently added: "If the author cannot be determined, then you may reconsider whether you have a reliable source at all." Was this on the basis of some consensus I missed? If so, I disagree. Official statements of organizations and governments, texts of laws, etc. are rarely attributable to authors; for that matter, articles in Wikipedia are rarely attributable to a particular author. I would really like to see this sentence either removed or expanded to the point where it resembles a sane policy. -- Jmabel | Talk 07:18, Dec 7, 2004 (UTC)

Dittoed. Johnleemk | Talk 07:39, 7 Dec 2004 (UTC)

Categories not Covered

I guess because I mostly focus on historical topics of Wikipedia I've encountered some problems in citation not discussed here. These are:

  • Citations from the Bible. The style I'm most familiar with is to write out in full the name of the book, then abbreviate it afterwards (e.g., Genesis & afterwards Gen.); chapter & verse are given in arabic numerals, separated by a colon (e.g. 3:16).
  • Citations from Classical sources. These are the ancient Greek & Roman authors: Homer, Hesiod, Vergil, Livy, etc. Whenever possible, write out the complete name of the work in Latin or Latinized Greek; italicize the title if it is considered book-sized, otherwise use quotation marks; book & line or section are given in arabic numerals, separated by a period (e.g. Iliad .100, & afterwards Il.; not Natural History but Naturalis Historiae 5.20, & afterwards N.H.)

ISTR that this is in line with the MLA guidelines, or matches established practice by scholarly classical journals -- but I could be wrong. Yes, I've been changing every citation I've seen to fit these; & yes, I've occasionally been inconsistent. -- llywrch 03:32, 16 Dec 2004 (UTC)

About the names of books, it seems like that would fall under the general rule of using whatever title the book is most commonly known by in the English-speaking world. Maurreen 05:15, 16 Dec 2004 (UTC)
One could finesse by making links to our growing collection of articles on the ancient source works, and making redirs from Latin/English as desired. Even after studying all this for some years, some of the names of works do not ring a bell for me (even when written out), casual readers will be even more mystified. Stan 17:17, 18 Dec 2004 (UTC)

What do you think about this? (footnotes and bibliography)

soren9580 and I have come up with a method of citing/referencing/footnoting pages that is very easy and very professional. Can we consider this method for the Wikipedia:Cite Sources page? Here is an example and we have created a page dedicated to this method as well, which serves as another example as well: Wikipedia:Footnote2. --[[User:Alterego|Alterego]] 22:58, 21 Dec 2004 (UTC)

Do I understand correctly that people have to hand-maintain keeping the numbers in the right order? -- Jmabel | Talk 23:30, Dec 21, 2004 (UTC)
Yes it's not that bad because they are very simple templates. We can hope that in the future there is a possibility of autonumbering. In the meantime, this is the only one that works and it is very simple. Just {{f1}} and {{f1b}}. Aside from this though, can we consider adopting this suggested Notes and Bibliography style for the encyclopedia? It looks very good I think. --[[User:Alterego|Alterego]] 23:43, 21 Dec 2004 (UTC)
Hand-maintaining is unfortunate, but I can't think of a better idea. This is less visually intrusive than the (author, year) in-line citations, and for texts with lots of disputes (say Internet Explorer) you really need in-line citations to justify statements. I think it'd make sense if you'd like! Maybe this will help gain experience, and then the software can be modified to support something like this but to number things automatically. Dwheeler 00:10, 2004 Dec 22 (UTC)
Without auto-numbering, I don't see a compelling reason to prefer this over in-text parenthetical citations. Bouncing around between anchor tags is, for me, a suboptimal reading experience, while parentheses can easily be skipped over by eye without web-browser assistance. So I'd prefer this to be presented as an alternative, not an all-Wikipedia prescribed style. Also, this introduces a potential new problem: many footnoted styles do not use a separate reference list. Do we intend to use both end-of-article notes and end-of-article reference lists? -- Rbellin 00:16, 22 Dec 2004 (UTC)
Hi Rbellin. Well, I don't think the point about in-line parenthetical references is good, because it's easier to skip over a tiny footnote if you don't want to read it. And it's also what we are used to seeing in books too. I would ask your opinion concerning the notes and bibliography section. I think it looks very nice in my example article, and in addition it is the proper way to do it when added with MLA styling. It is my opinion, and i know some people will find this extreme but that is ok, that since this is an encyclopdia even all of our external links should be in MLA style like i have done in MBTI. Just providing a link provides no context and no meta data and since the articles can be used under the GFDL outside of wikipedia, it seems very important to provide some information about the hyperlink that exists there. What do you guys think? --[[User:Alterego|Alterego]] 00:47, 22 Dec 2004 (UTC)
Yes, I use both footnotes and references (which the new suggested style suddenly changes the name of after we've been using the present style guide on thousands of articles), as far as Rbellin's question. I found this method useful so I could footnote my paraphrases but list references/bibliography for the other sources generally consulted for background information that were not paraphrased or quoted OR where a whole article owes a great deal to one source, listing the source in a References area is better than footnoting every single line. I think doing things with only a footnotes list would cause too many footnotes to be put in an article; having a separate references/bibliography leads to a cleaner look and fewer footnotes. I think footnotes should be just used occasionally when drawing particular attention to sources of a specific point, quotation or paraphrase that you can see might later lead an editor to want to verify it, but a References or Bibliography should always be used for the general background info. I guess I do like "Bibliography" better than "References" as a title. (I realize wikipedia used APA style for the "Notes" and "References" sections, having looked the other day. My old World Book Encyclopedia seems to only have "Additional Resources" as a section at the end of its articles and a byline for the article writer, but that is a different situation since our articles aren't written by solo authors credentialed in an area of expertise but are a joint (often anonymous) effort, and we have a need to verifiably show where we get our info so that readers and Wikipedia critics trust it.) I also used External Links to generally useful sites for more information for readers but which I likely didn't actually use information from in my article. One other thing: I also use footnote "short form" in re-listing article titles on the subsequent citation to a previously footnoted source, per most style guides, in case someone isn't used to that. Subsequent references to a source (short forms and ibid., etc.) could stand to be put in the footnotes part of the citation style guide. Emerman 08:36, 24 Dec 2004 (UTC)
In reference to your comments above about bibligraphies, as I said at Wikipedia_talk:Guide_to_layout#References vs Further Reading:
[for average users] it is important to distinguish between:
  • things they might find useful to go find and read, if they want more information about the topic than what is available here (what I am calling "Further reading"), and
  • the material used to create the article, which quite likely are detailed academic materials which would be of little use or interest to them (what I am calling "References").
... from my experience in exploring fields by going through bibiographies in books and getting items (something I have done in a very large number of fields) that the (sadly uncommon) bibliograhies which include comments about which items are good for what are a zillion times easier to use productively than the ones that just provide a barren list. I've lost count of the number of times I've ordered a book based on solely the listing in a bibliography, only to find out when it arrived that it was a waste of money. The article writers have (or should, if they are any good) an excellent understanding of which readings are best for "average people", and I think it's our duty to pass that very valuable information on.
I'm not so hung up on the particular titles we chose (although I admit a certain fondness for "Further reading" for "books the average reader should go look at if they want to know more than is in this article" - which I argue is probably important to more readers than the ones who want to know the exact source of each statement), so perhaps it's best if we first decide on how many sections there ought to be, and what the function of each is; we can worry about the exact names later. Eventually I'd like to see information on exact sources of article content moved to meta-data, but I think that's clearly a separate category of stuff than "Additional Resources" (also a choice I'd be happy with, although "Further reading" seems to have a large foothold already here). Noel (talk) 16:45, 27 Dec 2004 (UTC)
Generally concur. It bothers me that people keep collapsing "Further Reading" into "References": often I can say pretty confidently that the "Further Reading" has not really been mined for the article. -- Jmabel | Talk 19:05, Dec 27, 2004 (UTC)
Oh, maybe I should call "Further reading" certain things I called "External links"? How would you distinguish the purpose of an "External links" section from that? I will go back and look in our faqs for more info too. Emerman 16:17, 28 Dec 2004 (UTC)
This topic has been discussed at length here (see #References vs. External Links) and also at Wikipedia_talk:Guide_to_layout#References vs external links again. Noel (talk) 20:39, 28 Dec 2004 (UTC)
"External links" is probably OK for online material (although I'd love us to move to a standard where references are identified as References regardless of whether they are paper or online) but I was referring to print material. -- Jmabel | Talk 18:53, Dec 28, 2004 (UTC)
I think this would all be a lot less troublesome if we came to understand that there are two things which are logically separate axes: i) information used to produce an article, and ii) sources of further information for those who want to know more. Yes, the often overlap, but one is usually not a subset of the other. If we'd stop trying to cram things from these two distinct "meta-categories" into a single set of categories (created by dividing up one axis into categories) we'd have a lot less problems, I reckon!
Along the "further information" axis, I reckon it's useful to distinguish between "external links" (meaning other web sites on the topic, with all the plusses and minusses of the web - see previous comment), and "further reading" (altough some other name, such as "additional resources" would be equally OK with me). Making distinction is less useful when it comes to "article sources", and you have to do other things too - e.g. when referencing a web site (as opposed to directing users there for more info) it's important to specify which version (i.e. when) of the site you used. Noel (talk) 20:39, 28 Dec 2004 (UTC)
My last sentence above leads me to look at Chicago Manual footnote style and notice that they don't use "page" before page numbers (though I assume they would show a paragraph symbol or para. before a paragraph symbol in an online source) unlike the APA style I had found online. Chicago style (as well as Turabian) also doesn't repeat the article title the second time even partly, unlike another style guide I had referred to (but can't find now); they just repeat the author last name and page number, so I may leave out the article title in the short form. Emerman 16:10, 25 Dec 2004 (UTC)
My opinion: I like your handling of external links in the MBTI article (with extra information like date cited and author when known). I do not like the double full listings of reference info in the endnotes and in the reference list; if the reference list contains the complete citation information, the endnotes should probably just be relatively minimal pointers to the full citation, plus page numbers. About ease of reading, I simply disagree with you. And, as a sidenote: I have read lots of citation-dense books that do not use endnotes like these. They use parenthetical citations (or true footnotes on the same page, or marginal notes), two formats which allow easier reference without leafing from page to page (comparable to following anchor links to the bottom of the article and then clumsily trying to scroll back). In web-browsable format, probably the ultimate best solution would be marginal notes on the right in a wider browser window (a format that Edward Tufte advocates in his information-design books). But for now, I personally prefer parenthetical citation (for ease of maintenance and readability) over this solution, so I'd be against mandating it throughout Wikipedia. -- Rbellin 02:06, 22 Dec 2004 (UTC)
I just had an idea too. It would be really easy for someone to write a script that was onsite or offsite, where you put in the article coding and it went through and numbered everything for you. Possibly! --[[User:Alterego|Alterego]] 00:51, 22 Dec 2004 (UTC)
I spent a lot of time using the current style guide as well as info from the APA to do my Notes and References sections in the Kristin Hersh article and have seen others like this. Do I now change all articles using the old way to say Bibliography and find whatever things you've done differently from the previous way to copyedit things? Or is your example similar to the current style we already have? I suppose if the main change is to call what was once "References as now "Bibliography," that is not so hard; what are the main changes I'd be making to do this your way if it becomes adopted? I don't know if I can keep up with all future changes to the guide but I'll try. It seems like a style guide should have been developed prior to launching Wikipedia rather than have continual changes that will lead to inconsistencies between older stories referring to the old guide and newer stories where the writers referred to whatever was the latest version of the wiki citation style guide at the moment. The main point of a style guide is to create consistency, so I think a style needs to be developed and stuck by without changing it at some point. I do like how you just put the number to the note without "Note 1" or "Note 2" to start a citation though since I think that is the normal way to do it everywhere but Wikipedia, though I haven't double checked (is it APA style to say "Note 1" etc. before a footnote citation? If so then, I was mistaken). Also, I'm about to add a section at the bottom about how I cited to online publications without page numbering. Emerman 08:04, 24 Dec 2004 (UTC)

As I've explained when this topic was raised (by the same people?) in the past, splitting "References" into "Further reading" and "Bibliography" is a bad idea — there are good reasons why this is not done in scholarly articles. The main reason is that the distinction is completely artificial...especially in a multi-author work like Wikipedia, there's no clear distinction between works that were used to write an article, works used to verify it, and works suggested for further reading. (What you really seem to want are "advanced references" and "introductory references" sections. Of course, if there is a very long list of references, which is rare, there's no reason not to add a subsection heading or two.) The main thing is to cite readable references in the first place: Wikipedia is not for original research, and so it is preferred that we cite things like textbooks and review articles rather than scads of primary sources. —Steven G. Johnson 18:40, Dec 28, 2004 (UTC)

And I will reply (again) that Wikipedia is not a "scholarly article", and it would therefore be inappropriate to use scholarly article procedures here. The average encyclopaedia reader is not a scholar. I do wish people would put more importance on this point. Noel (talk) 20:39, 28 Dec 2004 (UTC)
I have never advocated blind adherence to procedures for scholarly journal articles in Wikipedia; please don't put words in my mouth. However, many of these procedures have been developed for a reason and have withstood the test of time, and it is wise to examine those reasons before disregarding them. (Moreover, referencing procedures are similar across wide varieties of publications; it is misleading to suggest that they are only used in abstruse technical articles with no relevance here.) —Steven G. Johnson
Sorry; I clearly didn't understand your "there are good reasons why this is not done in scholarly articles" comment. Anyway, I am very interested in importing good ideas from scholarly practise - e.g I'm one of the people who would like to see us have better citations for material in articles. In this particular case, however, it's the general practise of encyclopaedias (with their "further information") which I would like to follow. To borrow your words, "many of these procedures have been developed for a reason and have withstood the test of time, and it is wise to examine those reasons before disregarding them." Noel (talk) 13:52, 29 Dec 2004 (UTC)
Also, as I pointed out above, we probably do need to distinguish between things used to write the article, and suggestions for additional resources, because I think a lack of this distinction is behind a lot of our problems here. Noel (talk) 20:39, 28 Dec 2004 (UTC)
The point is that you cannot in general distinguish the two, however much you might like to, and attempting to do so artificially leads to even more confusion about citations. —Steven G. Johnson
I also have to disagree very slightly with your point about 'cite readable references'. I do agree whole-heartedly that it's a absolutely desirable goal, and one that should be ignored only if no other course is possible. However, there are circumstances in which a reference to a not-widely-available work is necessary, especially when making a point that goes against the "received" wisdom. (E.g. I just had reason today to add a reference to a complex scholarly article to HMS Hood (51) because it's more recent work which debunks the common wisdom that it was weak deck armour that doomed her.) Sometimes, alas, in persuit of our goal to make Wikipedia a source that people can trust, there's just no substitute for a reference to a primary source which is not, alas, suitable for the average reader. (Again, refernces != further info....) Noel (talk) 20:39, 28 Dec 2004 (UTC)
Certainly, one sometimes cannot avoid citing a range of sources, from layperson to specialist. In most cases, however, a reader is perfectly capable of distinguishing the difficulty of the sources simply by the title (e.g. is it from a newspaper, or a textbook, or a scholarly journal?). In the worst case, there's nothing wrong with adding a comment to the citation information in the references. —Steven G. Johnson 02:33, Dec 29, 2004 (UTC)
Alas, I wish I could be so sanguine! Yes, if we could be certain that editors would add illuminating comments, I would be happy to go along with your proposal. However, I fear that most will not take the time. (Heck, most don't even take the time to list sources/etc!)
And as for being able to do it from the titles, there I must emphatically disagree with you. The shelves of my library are littered with books which I bought from the title, having seen them listed in an un-adorned bibliography, and which turned out to be worthless. Noel (talk) 13:52, 29 Dec 2004 (UTC)

Regarding numbered styles, I think it is a bad idea until/unless MediaWiki has autonumbering support in software. (The parenthetical style is also easier for new users since they don't have to learn yet another wiki syntax.) —Steven G. Johnson 18:40, Dec 28, 2004 (UTC)

Linking authors

I don't like the recent change that links a bunch of author names and at least one word in a title. I think this distracts from the main thrust of this project page. I'd rather revert that and just have one paragraph on the fact that links from within citations are allowed, where appropriate. Does anyone strongly disagree? -- Jmabel | Talk 23:26, Dec 22, 2004 (UTC)

Sounds reasonable to me. —Steven G. Johnson 01:27, Dec 23, 2004 (UTC)

Citing to online publications in footnotes by paragraph number

I recently saw someone misunderstand why I had paragraph symbols in my footnotes to Kristin Hersh because I don't think they had seen the current Cite sources wikipedia guide mention it and the person thought I meant to use page numbers till I explained it to him. I had found the paragraph symbol method by going to the bottom of the Cite Sources page to the link to Citation Style Guides, and clicking the link at the top for APA style. That page includes the following example:

For electronic sources that do not provide page numbers, use the paragraph number, if available, preceded by the ? symbol or abbreviation para. If neither is visible, cite the heading and the number of the paragraph following it to direct the reader to the quoted material.

(Myers, 2000, ? 5)

(Beutler, 2000, Conclusion section, para. 1)

Examples addressing online publications without page numbers should be in our wiki cite sources page because people need to be as specific as possible when citing in a footnote to the source of material we directly paraphrase or quote.

Here is MLA style on the same subject:

If your source includes fixed page numbers or section numbering (such as numbering of paragraphs), cite the relevant numbers. For numbers other than page numbers, give the appropriate abbreviation before the numbers: "(Moulthrop, pars. 19-20)." (Pars. is the abbreviation for paragraphs. Common abbreviations are listed in the MLA Handbook, sec. 7.4.) Do not count paragraphs yourself if your source lacks numbering.

So, MLA would have me not cite to a paragraph of an unnumbered webpage. That makes footnoting to an online periodical less specific in showing people the location paraphrased material was taken from. I am curious as to whether I should go with APA or MLA since there is a discussion about this here. I have used APA so far in footnoting paraphrases from such material so far (mainly because that's what I thought the Cite sources page recommended), and I think it's actually better to provide the specific paragraph number when paraphrasing or quoting. (Of course it's easier not to. And here's another source that tends not to get into paragraph numbers: Columbia Online Style). Emerman 08:13, 24 Dec 2004 (UTC)

Citing Other-language wikipedias

I've started a new section on citing articles from other-language Wikipedias. I won't be surprised if we have to duke it out a little to get a standard, but up to now there has been none. I do a lot of translation, and up until now, I find that in the absence of a standard, whatever I do people want to change it (or, worse yet, just delete it). We need a standard for this, even if it turns out not to be the one I am initially proposing. -- Jmabel | Talk 19:24, Dec 27, 2004 (UTC)

References, sources, etc.

People are using "References" and related sections differently, some just as "Further reading," etc. I'd like to suggest "Sources" for material used to write or verify the article. I think that word is more clear. Maurreen 19:51, 28 Dec 2004 (UTC)

I've been using "References" for "things used to write the article which are not recommended to the general reader", but I would be happy to switch to "Sources" or any other similar term which is agreed upon for that purpose. Noel (talk) 20:53, 28 Dec 2004 (UTC)
Having separate sections for further reading vs. sources used to write/verify an article is a bad idea, as there is no clear distinction between the two (especially for articles with multiple authors). This has been discussed multiple times before; please see the previous discussions. —Steven G. Johnson 02:17, Dec 29, 2004 (UTC)
The problem, from my point of view, is that in many articles (certainly this is true of most of the ones I write), of the two categories, neither is a subset of the other. So one can't subsume one into the other. Short of creating a name for the intersection set (gag!), I don't see any easy answer other than repetition. Yes, I know that's ugly, but the alternative (having only one list) is one I like less. Noel (talk) 13:38, 29 Dec 2004 (UTC)
It's been reiterated, but it hasn't really been discussed. I know that I sometimes add, as "further reading", a book for which I've only read a review, but haven't even seen the book, let alone used it as a reference. -- Jmabel | Talk 06:30, Dec 29, 2004 (UTC)
Your point is what, exactly? Yes, you should ideally only cite things that you have looked at yourself, but this is equally true for "further reading" sections. Yes, often in the real world people just want to point readers to a "standard" textbook/article on a subject, and they cite the standard book (e.g. one recommended by a colleague) without looking it up themselves...and this happens for "references" too. It seems to me that you are begging the question, because your apparent point is predicated on the assumption that "references" are only things that are used to write an article — this is true in school assignments, maybe, but not here or many other "real world" cases. —Steven G. Johnson 07:19, Dec 29, 2004 (UTC)

Put in a more practical terms, suppose Wikipedia recommends separate "Sources for the article" and "Further reading" sections. Let's ignore for a moment the endless confusion that will result because many references will fall immediately into both categories. Consider instead what happens when someone else comes along, looks up a "further reading" book, and uses it to check an item in the article. Should they move the reference into the "sources" section? Or what if someone thinks that a "source" makes good further reading? Should they move it? Please, let's avoid a nightmare caused by an ill-defined breakdown. —Steven G. Johnson 07:19, Dec 29, 2004 (UTC)

Sigh. If only we had citations as real meta-data, a lot of these problems might go away... Noel (talk) 13:38, 29 Dec 2004 (UTC)

(If you come across some rare article with dozens or hundreds of references, by all means break down the reference list into subsections such as "introductory" and "specialized", or whatever would be helpful to a reader. I'm not opposed to organization, in the rare cases where it is needed; I'm only opposed to a universal attempt to impose a false dichotomy on citations. —Steven G. Johnson)

You ask (probably rhetorically), '...what happens when someone else comes along, looks up a "further reading" book, and uses it to check an item in the article. Should they move the reference into the "sources" section?' I'd answer (non-rhetorically) that if that fact deserves specific citation, then they should make the citation clear and add it to the references.
I have no objection, though, to lumping far more of these things together: in practice, subtle distinctions are unlikely to be maintained; I also find the sharp distinction between print and online sources artificial, especially since so many things are now both. When I encounter those, I try to give the full reference for original print publication, but also provide an external link. -- Jmabel | Talk 07:53, Dec 29, 2004 (UTC)
Yes, I do that too. Specific documents which I used as "sources" (in the sense you've been using) which are available online I put in the "References" section, and link to them - because generally such things don't have the "which version of the web site are we talking about" problem. I put web sites on the topic (where the content may change) into "External links". Check out e.g. HMS Hood (51) for an example. Noel (talk) 13:38, 29 Dec 2004 (UTC)

What Maurreen is proposing may be "ill", but it is "definition" and I find it laughable that her suggestion would lead to a "nightmare" situation (unless one has a phobia of overinformation). I would assume that if people keep bringing this up there is already a problem. Perhaps the relevant policy needs to clearly explain Steven's point so that it is not brought up as often: Obviously Maurreen has seen the citation policy and it did not lead her to the same conclusion as Steven. Hyacinth 19:02, 29 Dec 2004 (UTC)

Adding that clarification seems like the best solution to me, at least for the short term and possibly for good. How about this? -- Rbellin 23:58, 29 Dec 2004 (UTC)
The References section should list both recommended further reading for interested parties and sources cited in the article. Adding known useful readings to an article's reference list makes it much easier for later editors to confirm the article and add citations for further information, even if you don't have the time.

There is a proposal for a separate namespace for sources, see bug 1199.

My opinion is that it would be preferrable to have "further reading" and "external links" in articles, and keep sources out. Source information should be kept out of articles for the same reason that discussion is kept on talk pages. Consider for example how currently WWW references are (recommended to be) given with the date for their retrieval, which certainly is relevant to editors but irrelevant to the article subject.

There are many advantages to this scheme. The most obvious is that one can provide detailed references, page and line numbers if necessary, for each fact — without kludging up the articles. Another advantage appears when the most authoritative reference for a fact is located in a printed work that may be difficult to access for most Wikipedians for verification purposes. Instead of just listing this work, which is the only reasonable way inside an article, one can on the sources page list 10 websites that contain the same fact, so editors can verify it quickly. Discussion of source material and directives for other editors can also be included.

Duplication between source lists and further reading lists seems like a minor issue to me. Fredrik | talk 00:23, 30 Dec 2004 (UTC)

Using just a "References" section and no others is OK with me. My original point was that the current common division between "References", "Further reading", and "External links" is ambiguous. I agree with Jmabel that having Web sites or pages in a separate section is an artificial distinction. Maurreen 05:07, 30 Dec 2004 (UTC)
It all depends on what kind of web reference it is, I think. If it's to a particular fixed document (e.g. a magazine article, or published paper) that someone has made available in its entirety online, it's quite proper (IMO) to list it in whatever category you feel it belongs in. It's OK to list as a source/citation, because even if the web site goes away, the link was never more than a convenience anyway - the original document is still acessible (albeit a lot less easily).
If, on the other hand, if the link is to an entire site, the content of which is changing often (as many sites do), and it's linked to as a source of more information, we have to recognize that what's there in the future, when some reader clicks on the link, may not be what's there now (if the site is there at all), so it really is in a different category from a book which one would list in "Further reading". (Although a link to a book which has been made available online would be fine in "Further reading", of course.) Noel (talk) 14:31, 30 Dec 2004 (UTC)

Stevenj's recent edits

I strongly disagree with Stevenj's recent edits drastically reducing the discussion on how to correctly cite books, and assuming that it's just a matter of telling people to use templates.

  1. Somewhere we need to discuss the actual policy on what is correct. I think this is the place to do this. Stevenj's edit is tanatamount to burning the requirements document because you believe you now have a correct implementation... or at least one that can be tweaked.
  2. Furthermore, not everyone likes using templates. Someone should be able to look at this page and see how to write a citation. They should not be effectively told, "hey, if you are not comfortable with parameterized templates, don't try to cite things."
  3. Even if you disagree with me on the previous points, this change is incredibly premature. We have effectively dropped discussion on citing a particular article within a larger book and on handling the situation where a book is attributed to particular editors rather than particular authors.

It is possible that some of the deleted material is, indeed, redundant, but I stand by my general points.

I don't want an edit war, so I will not revert; I would appreciate other people's comments on this. -- Jmabel | Talk 22:53, Jan 1, 2005 (UTC)

I strongly agree and feel the examples should be immediately replaced. There is no consensus at all that the templates are preferable to plain-text citations. (And even if there were such a consensus, it would behoove us to support non-technically-minded users with strong examples which provide all the information necessary for later template cleanup.) The removal of the discussion of the citation templates' technical limitations is especially strange; has the cap on 5 citations per template been removed (and in that case, why preserve "template 2" and "template 3"), or are we just brushing it under the rug? -- Rbellin|Talk 23:06, 1 Jan 2005 (UTC)
Yes, MediaWiki 1.4 (we just switched to it) removed the "only allowed to use any template 5 times per page" limit. I imagine we'll have to keep template 2/3 around until all the pages that use them get updated. (Although I suppose they could be turned into redirects.) Noel (talk) 23:11, 1 Jan 2005 (UTC)

I am a little baffled by your objections, since it seems we all agree that templates are not necessarily the way to go. I totally agree that plain-text citations may well be preferable. (My removal of how to cite an article in an edited book was accidental; I've reinstated it.) Since User:Pcb21 added a whole section on the use of templates, I figured that it was redundant to put the same information in the "Citing a book" section so I removed it from the latter. Especially given that it is controversial, I think the template stuff should all go into one section (especially to minimize the confusion to new users). —Steven G. Johnson 23:35, Jan 1, 2005 (UTC)

That looks better! Thanks for fixing this so quickly. I agree, it's probably better to put all the template information together, and if possible in a form that parallels the plain-text citation examples so that comparison is easier. -- Rbellin|Talk 23:43, 1 Jan 2005 (UTC)

Templates vs. Plaintext References

Disregarding the comments arising from accidental/interim changes, the important point arising here is

Why are templates not the way to go?

Jmabel's objection is "some people don't like using templates". Could you expand upon this please? Clearly we will need more than one template. To get a usable set, would we have too many templates? I don't think so. But let's discuss that. If the idea of having a whole references tab, or perhaps a metadata tab continues to gain momentum, having these templates around are going to be a tremendous boon. If there are objections, I want to know about them and to know whether they have merit. Pcb21| Pete 00:04, 2 Jan 2005 (UTC)

My concern is simply that the template adds yet another syntax that new editors must learn in order to add reference information, which may deter them from adding references at all. Templates address the relatively unimportant problem of formatting while (perhaps) worsening the primary problem that people don't reference in the first place. (Yes, editors can still add plain-text references that get templatized later, but if the template becomes widespread I worry that a new editor will open the "References" section and get put off by all of the non-plaintext syntax.) Let's not lose sight of the forest here. —Steven G. Johnson 00:32, Jan 2, 2005 (UTC)
Sure absolutely, the most important thing is getting references however they come. However, even "plaintext" is not really plain text - new users learning the syntax by copying other pages will still see a sea of markup for bolding and italicising. My feeling was that the template version of a reference was actually more readable and intuitive than where you have to specify the formatting yourself. But perhaps my view is coloured by a) my using BibTeX all the time and b) my being familar with templates. Pcb21| Pete 11:34, 2 Jan 2005 (UTC)
As another issue, plaintext is much easier to modify for special cases than templates, if a particular citation doesn't fall neatly into a particular cubbyhole. If you look at a really serious citation formatting system like bibTeX you'll see it has an enormous number of optional fields to cover all of the various cases (and these still aren't quite enough...errata aren't handled gracefully, for example). One can use a mixture of plaintext and templates, of course, but then you lose most of the advantage of templates (the ability to reformat all references on the fly, use "metadata tabs" etcetera.) —Steven G. Johnson
Yes my long-term hope is that a BibTex file would be readable by the MediaWiki software and it would spit out appropriate HTML (likely on the metadata tab page). Perhaps we would then end up with one .bib for each WikiProject or something like that. Power users would then take advantage of the powerful BibTeX system. Regular users would carry adding references in plain text. However that is clearly sometime away (and if it happens at all it is because some developer is interested in it) but I thought an appropriate suite of templates would take us along the path.
I don't use templates. This doesn't mean I object to anyone else using them for this purpose. But so far, they aren't worth the trouble for me. I'm not saying anyone should cater to me, but I doubt I'm a minority of one. Maurreen 05:24, 2 Jan 2005 (UTC)
Ok, knowing that our policy has to cater for this point of view, it is clear we cannot mandate a template-only approach at this time. I understand where you are coming from when you say "so far, they aren't worth the trouble" if you are talking about templates in general - mostly they are used for metadata junk rather than to add much to the article. Have you seen how they are used in practice for references though - my feeling is that the formatting using Template:Book reference is actually neater and more intuitive than direct formatting. Have you tried that? Do you disagree? Pcb21| Pete 11:34, 2 Jan 2005 (UTC)

I took a look at Template:Book reference, and I don't see the value. But if other people want to use it, that's fine with me. Maurreen 18:48, 2 Jan 2005 (UTC)


I didn't think we had anything like consensus in favor of Template:Cite sources, so why is it mentioned on this project page (without even any discussion of the arguments as to why it may not be useful)? -- Jmabel | Talk 06:37, Jan 4, 2005 (UTC)

It doesn't appear to be discussed on this page at all ... or did I read too quickly. Incidentally, I am firmly not in favour of having the template dominate the whole article appear at the top. The talk page is the most appropriate place for it. Pcb21| Pete 18:18, 4 Jan 2005 (UTC)

I've removed the following from the article until we can achieve at least a basic consensus that (a) it is useful and (b) where such a tag should be placed. —Steven G. Johnson

If you want to mark an article as having inadequate citations, you can do this by including {{Cite sources}} at the top of the text — this inserts a template notifying people that the article needs citations, and automatically includes the article in the category for missing citations. Note: the appropriate use and phrasing of this template is still under discussion. Please see Wikipedia talk:Cite sources and Template talk:Cite sources.

I personally think that a template of some sort is useful (particularly in cases where you don't have a specific reason to doubt a claim, but are uneasy because you have no way to verify it). Unlike Pcb21, however, I believe that it clearly belongs on the disputed article itself as for {{NPOV}} and {{disputed}} and {{dubious}} and {{stub}}, rather than in Talk. Like for the latter tags, putting it on the article serves three purposes: (a) warning readers that a specific article's reliability has been found difficult to verify (beyond the general disclaimer that applies to all Wikipedia); (b) making sure editors see it (since a notice can easily be buried in a long Talk page, like this one); and (c) serving as a barb to spur editors to fix the article. —Steven G. Johnson 19:51, Jan 4, 2005 (UTC)

I don't have a strong feeling either way about whether it belongs on the article page or the talk: page - either is fine with me. I do note that there has been a recent campaign to move most of the messages concerned page quality to the talk pages, and I would assume that the people doing that would consider this message to be similar.
I think you have a good point about it being buried, so if it does go on the talk: page, it should go at the very top, where it's more likely to be seen. Noel (talk) 21:38, 4 Jan 2005 (UTC)
I think editors are likely to miss it even at the top of a talk page, since the usual habit is to scroll to the bottom. This is the first I've heard of a "campaign" to move dispute/complaint tags to Talk; can you provide a link? —Steven G. Johnson 22:00, Jan 4, 2005 (UTC)

I can't even remember where this was previously discussed (maybe the Village Pump?) but my concern remains that merely saying that an article lacks citations isn't very useful. Unless we were to demand that all articles meet an academic level of citation, citing even for things that are common knowledge -- which I simply don't see happening in the foreseeable future -- citation issues are always going to require specification of what specific matters are doubtful and require citation. If we can agree that this is parallel to the NPOV tag or the dispute tag -- not to be used lightly, and requiring an explanation on the talk page of exactly what issues are disputed (or, in this case in need of citation), then I'm all for it, and don't care which page it is on. If it is not accompanied by such a requirement, I see it as a sheer liability. -- Jmabel | Talk 04:36, Jan 5, 2005 (UTC)

Prior discussion from Village Pump

I found the prior discussion from the Village Pump. Here it is. -- Jmabel | Talk 05:15, Jan 5, 2005 (UTC)

Moved from [Wikipedia:Village pump (proposals)]]: CiteSources template and Missing Citations category

I've created a new "CiteSources" template to help identify articles that don't cite their sources. If you come across (or write!) an article that doesn't adequately cite its sources, you can add {{CiteSources}} to the top of the article. This will automatically add it to the Category:Missing Citations, which maintains a list of the articles missing citations.

Nobody's obligating anyone to use it. But, if you'd like to use it, here it is! -- Dwheeler 01:00, 2004 Dec 29 (UTC)

While I am a very strong advocate of citing sources...
  1. Wouldn't something like this be better on the talk page?
  2. Citing sources is not usually a black-and-white matter. For example, it's really no problem if we don't cite a source for the readily ascertained date of the JFK assassination, but it's very bad to have an unattributed claim about gunmen on the Grassy Knoll. This kind of thing isn't easily dealt with by slapping on a template. I think, like the NPOV template, this should always be accompanied by a specific indication of what facts need citation, so it can be removed when the matter has been addressed. -- Jmabel | Talk 06:21, Dec 29, 2004 (UTC)
Further, while it is worthwhile to cite sources for "news/event reporting" facts and "opinions of experts and opposing forces" facts, the kinds of facts that are generally well-understood or are very uncontroversial wouldn't seem to need cites in the article itself (but perhaps discussed out in 'talk' instead). Also, consider that many descriptions, if not written in a fully factual manner or don't provide full coverage, will be edited over and over again until they meet the community's agreement for NPOV and factuality. --Stevietheman 06:42, 29 Dec 2004 (UTC)
Such a template should go on the talk page and not on the article itself, IMO. olderwiser 14:49, Dec 29, 2004 (UTC)
If you think it's better on the talk page for a particular article, you can obviously put it there. I'm not even saying anyone has to use this at all; I just thought it'd be a useful tool for those who want to use it. However, I think it would usually be better to put this template on the article itself, and not the talk page, for the same reasons we put the NPOV and stub templates on the articles themselves. By placing this template on the article, readers of that article are given a heads-up about a known problem and are invited to help fix it. I think that most of the time people will view articles, not their talk pages. If you only place the template on the talk page, many people you want to notify won't get the notification. NPOV and citing sources are two of the standard ideals for articles, and sometimes it's hard to deal with POV without the sources being identified, so it makes sense to give them similar treatment. Early in Wikipedia's life, the question was whether or not there'd be enough people to get any information at all. I think that question has been resoundingly answered. Now the question is whether or not the quality and reliability of the articles will be sufficient for it to be useful. I think citations are important to getting there, so little tools like this can help people know where they are missing so the citations can be added. -- Dwheeler 19:37, 2004 Dec 29 (UTC)
There are some meta-templates that are appropriate to appear on the article (such as VfD, NPOV, or disambig) and some that are less appropriate, but widely established (such as stub). However, in general there should be an extremely compelling reason to clutter up an article page with meta-comments. NPOV and related tags are important to place on the article page to alert readers that the content of the article is contested--it functions to make casual readers aware that the article may contain factual inaccuracies or biased statements and to use the content carefully. I have a few objections to placing the CiteSources tag in the article: 1) it is IMO rather subjective, since a majority of articles in Wikipedia could arguably be tagged as not citing sources sufficiently. 2) The tag in isolation is unspecific as to in what way the article is lacking citations--is it a specific fact or in general? Regardless of whether the tag appears on the article or the talk page, there should be a justification for the tag on the talk page explaining exactly what areas are felt to be lacking. It is similar to expectations for adding an NPOV tag--there should be some explantion for the tag on the talk page--and without such a justification, the tag should certainly not be allowed to remain on the article page. 3) Having the tag appear on the article provides very little benefit to the casual reader--most casual readers are likely not interested in doing research to provide citations for unspecific reasons (and most would likely not have sufficient resources or interest to do research for specific reasons). If the lack of citations is sufficient to cause one to doubt the veracity of the article, then a different sort of tag should be used to alert the reader. However, if you're not alleging the possibility of factual errors or POV biases, but simply requesting assistance to improve the article, it seems that is more appropriate on the talk page, similar to how the various tags on Wikipedia:Cleanup are supposed to be placed on the talk page and not on the article. Please understand, I am supportive of the need for better citations in articles. However, I suggest incorporating this template with the existing set of cleanup tags. olderwiser 18:18, Dec 30, 2004 (UTC)
I was modeling this after NPOV and stub, which do appear in articles. In fact, this is the first time I've heard that placing "stub" on an article (instead of its talk page) was controversial to anyone. I took a look at the talk page for cleanup, as you suggested. While placing this on talk pages is certainly suggested here, it's not clear to me that there's a consensus that this info really should be on a talk page. In fact, under the "Talk?" heading in the talk page of Wikipedia:Cleanup, I see that MacGyverMagic, Icundell, and CDN99 strongly opposed placing the notices in talk pages and wanted them in the actual articles instead. rbrwr was ambivalent in general. Nobody seemed to support the notion that these notices should be in the talk page, although there must be at least one person who does since the text suggests it. If that's representative, perhaps all cleanup notices belong on the article itself, and we should fix Wikipedia:Cleanup to reflect that? That's probably a pointless exercise in argument, though. A simpler approach might simply be to leave it to the discretion of the submitter -- no matter what, this tag should be temporary anyway! Thus, if the person adding this mark thinks it's important for readers to know about an absence of citations, then put it in the article, else in the talk page. Since these markers go away as the information is added, it's not clear that a strong standard is really needed. -- Dwheeler 21:41, 2004 Dec 30 (UTC)
Now that the template is there to experiment with, it's interesting to see what people are doing with it. It's been added to Semen and Remote Viewing by others (not me), and in both cases to the article not the talk page. (Warning: Semen seems to be a high-vandalism article.) I presume they find it okay to put it in the article itself, so I must not be the only one who thinks that'd make sense. The Semen article uses a third option, interestingly enough, not discussed so far. In the Semen article, the template is used in the article itself (not its talk page), but instead of placing the template at the top, it's placed where the references would go if there were any. Perhaps suggesting placing this template where the citations are or would go be a reasonable compromise for this case...? -- Dwheeler 23:39, 2004 Dec 30 (UTC)
By the way, I disagree with the idea that "obvious" information never needs a citation. Certainly our current guidelines for ideal articles say that all articles should have citations. Things that are "obvious" are often less so over time, and over time some information gets progressively harder to find. Sometimes, in the process of finding our citations, we discover that the assertions aren't right! I don't view lack of citations as a "cleanup" anyway; grammatical problems and the like can often be fixed by lots of people, in many cases without even in-depth knowledge of the subject. But finding good citations can often be really hard for anyone other than the original author (where did you get that claim about XYZ?) And in any case, articles really need citations to give (1) readers a reason to believe in the quality of the article, and (2) helping readers/fact-checkers verify the assertions in the article. Anyway, it doesn't sounds like we're arguing over the goodness of citations, but merely where to best place a note when they're weak. -- Dwheeler 23:50, 2004 Dec 30 (UTC)
But finding good citations can often be really hard for anyone other than the original author, which is precisely why the tag would be more appropriate on the talk page rather than cluttering up the article with vague suggestions that the article might be incorrect without coming right out and saying so. There is a big difference between suggesting that an article is inaccurate or is biased and identifying that an article could use some citations. I mean seriously, most articles in Wikipedia could arguably use more citations--so what? Unless there is some specific indication about what statements could use citations, the tag really isn't very helpful, and especially not for the casual reader. For individuals with a passion for researching, having a tag on the talk page and a category should be sufficient to help in identifying articles that have been marked as needing attention. I see no need to place vague insinuations about the quality of the article within the article itself. olderwiser 01:25, Dec 31, 2004 (UTC)
Concur. Strongly. -- Jmabel | Talk 03:10, Dec 31, 2004 (UTC)

<end moved content>

Making Template:Book reference and Template:Book reference 2 compatible

Template:Book reference (hereafter, T:BR)and Template:Book reference 2 (T:BR2) are very similar in purpose, except for T:BR2 being for situations when the range of pages needs to be included in the citation. User:Rdsmith4 attempted today to alter T:BR to match the other, but I think some more discussion of what's actually the right thing to do is called for.

Differences / suggested changes:

  • T:BR includes the '*' for the list item, while T:BR2 does not. I think the latter is the correct behavior, because it is the more intuitive, because it allows use in other context than a bulleted list, and because it breaks the concept some of the developers have advocated, that templates are self-contained rather than depending on syntax in the containing documents.
  • T:BR2 is hard-coded to insert the 'ISBN', and thus works only for books that have ISBNs. This excludes many older works. I think that T:BR's use of an 'ID' field works better; it allows any cataloging number.
  • T:BR has capitalised argument names, while T:BR2 does not. Which is the preferred style?

Anything else? Any opinions? —Morven 08:38, Jan 5, 2005 (UTC)

Just in case you were wondering BR and BR_n were implemented independently and then later had only their names merged whilst the slightly more onerous task of uniformizing the input was left to, umm, now.
Re your points in order:
I prefer BR_n on this.
I prefer BR on this for the reasons you cite.
I don't have a preference but.... BR is used quite a bit (>100 pages). BR_n are not widely used. Removing the "*" from the BR would be quite quick to fix, but removing capital letters is a little fiddly. Maybe stick with BR for pragmatic reasons? Pcb21| Pete 11:59, 5 Jan 2005 (UTC)
In response to this, I have made the first two changes: T:BR now no longer includes the * for the list item, and T:BR2 now uses 'id' instead of 'isbn' so that non-ISBN books can be cited using it. I have edited all the articles that use these templates to correct their syntax.
I am still undecided about the argument capitalisation, but after correcting all the T:BR articles for the list item star, I don't think I want to go through that again, so that may be sufficient argument to fix T:BR2 etc. instead! —Morven 17:56, Jan 6, 2005 (UTC)