Wikipedia talk:Citing sources

Footnotes and punctuation

The para should be changed to one standard, the wiki and Chicago MOS standard of footnotes coming after punctuation, not both. There's some to be said for consistency. The option to have footnotes before punctuation should be stricken. — Rlevse • Talk • 19:16, 19 February 2008 (UTC)

The current language was the result of a recent lengthy dust-up on that issue. Personally I think that we might as well have editors act at their discretion in areas where consistency is of no discernable value. Christopher Parham (talk) 01:44, 20 February 2008 (UTC)

I think simply having consistency over either a large segment of Wikipedia, or over at least the more prominent articles, would be a discernable value. I'm not interested in revisiting that "lengthy dust-up", but I would like to ask a related question: is the Vancouver system of citation acceptable on Wikipedia, or not? Gimmetrow 02:13, 20 February 2008 (UTC)

It's of value if a reader likes consistency for its own sake, but it's not obvious to me why one might do so. Most readers will approach Wikipedia one article at a time and minor inconsistencies across the collection will not be troublesome. Christopher Parham (talk) 02:34, 20 February 2008 (UTC)

There are benefits to editors, too. But like I said, I'm not really interested in revisiting that. Now what about Vancouver? Gimmetrow 03:30, 20 February 2008 (UTC)

Agree with Rlevse and Gimmetrow; the last go-round wasn't so much a dustup as people just grew tired of the edit warring from a few editors. Wiki has reached a level of importance that warrants having our own housestyle, for consistency and maintenance purposes. I support a return to the long-standing, previous wording, which reflects consensus and practice on Wiki. SandyGeorgia (Talk) 03:52, 20 February 2008 (UTC)

The wording before the edit warring last October was:

Where to place ref tags

Place a ref tag at the end of the term, phrase, sentence, or paragraph to which the note refers.

The ref tag should be placed directly after most punctuation marks, without an intervening space in order to prevent the reference number wrapping to the next line. The same is true for successive ref tags. The exception is a dash—which should follow the ref tag. This is the format recommended by the Chicago Manual of Style.

Example:

According to scientists, the Sun is pretty big,<ref>Miller, E: "The Sun.", page 23. Academic Press, 2005</ref>
however the moon is not so big.<ref>Smith, R: "Size of the Moon", ''Scientific American'', 46(78):46</ref>

== Notes ==
<references/>

Comment: It might be useful to think of users as human beings, members of a species able to keep only a limited number of things in their heads at one time, particularly when their participation is as volunteers without renumeration. Which policy or guideline are we willing to put up with users forgetting about to make room for this? Feels WP:CREEPy. Best, --Shirahadasha (talk) 22:57, 20 February 2008 (UTC) Note: I'm not against easy-to-use tools that assist users in providing nice-looking, standardized styles or the creation of bots that can conform styles after-the fact. I'm against imposing these sorts of style requirements on the backs of ordinary users, particularly if not accompanied by any proposal for tools or bots to make the burden easier. Best, --Shirahadasha (talk) 23:01, 20 February 2008 (UTC)

Shira, this is not something brand new. It was the guideline for over a year, and there are tools to take care of it, so it imposed nothing on the backs of ordinary users. Gimmetrow 23:07, 20 February 2008 (UTC)

This is very helpful and very easy to understand. I don't know why it was removed and it needs to be reinstated asap. Tyrenius (talk) 02:34, 21 February 2008 (UTC)

Programmatically, writing a tool to "fix" misplaced references after punctuation would not be a terribly difficult task—assuming such tools are not already available, as it appears there might be—and should little affect the decision. The MOS has previously drawn hard boundaries on style considerations which vary outside of Wikipedia, for example, on punctuation inside of quotation marks. The guidelines should do the same here, without any pussyfooting. I recommend restoring the previously existing rule to a simple, easy-to-follow, requirement to put ref tags after punctuation as indicated. -- Michael Devore (talk) 01:01, 23 February 2008 (UTC)

"Some editors prefer the style of journals such as Nature, which place references before punctuation." - This wording, given in support of the option of punctuation before ref tags, seems rather a weak rationale to me. Now, if I were invited to contribute an article to Nature, then, sure, I'd follow their house style too. But as I contribute to Wikipedia then I'll try to follow Wikipedia's established style. The expectation for different editors to follow consistent style guidelines rather than all simply doing their own thing is really the whole the point of establishing style in the first place. I don't really see the need for guidelines to pander to one or two editors who can't get their heads around this. --SallyScot (talk) 10:42, 23 February 2008 (UTC)

Where were you all when this was dragging on for months? Gimmetrow 23:07, 23 February 2008 (UTC)

• Asking a question again for lack of response: is the Vancouver system of citation acceptable on Wikipedia, or not? Gimmetrow 07:09, 26 February 2008 (UTC)

Re Vancouver. According to Wikipedia:Citing_sources#Citation_styles it sounds like pretty much anything goes. Whether we might want to be more selective and venture to establish a smaller subset of preferred styles for Wikipedia is another question. --SallyScot (talk) 20:23, 29 February 2008 (UTC)

Yes, the CITE guideline is clear that any system is acceptable: "Any style or system is acceptable on Wikipedia so long as articles are internally consistent." This is an ENGVAR-type issue: everyone believes their own system is superior, so we generally just take the first established style unless the editors of an article agree to change. — Carl (CBM · talk) 23:49, 29 February 2008 (UTC)

---

Re. where were you all when this was dragging on for months? - From my initial readings on the subject I'd kind of assumed that the dispute had been going on forever. If on the other hand it's more true to say that the CMoS recommendation was in place well before one or two editors came along with their alternate style then I'd suggest they need to do better with their argument than "some editors prefer...". It would be like a new editor coming along now and insisting on capitalising all words in section headings (e.g. Where To Place Ref Tags instead of Where to place ref tags) on the basis of their preferring the style of some other publication.

Perhaps I've missed some salient argument, perhaps previous discussions became entrenched and adversarial instead of encouraging consideration of alternate rationales. In any case I invite proponents to summarise their reasoning again here.

I must say however, if all that can be vouched for the approach of putting references before punctuation is "some editors prefer" it, then I'll suggest there as no serious objections to restoring the CMoS recommendation that references should follow punctuation - on the basis it is the style which was established here first.

--SallyScot (talk) 12:50, 3 March 2008 (UTC)

I suggest that you look through the archives of this talk page and that of footnotes. If you do you will see that this dispute had gone on for more than a year. I strongly suggest that the compromise is left in place. No it is not true that "the CMoS recommendation was in place well before one or two editors came along with their alternate style". I can provide links to archived sections links dates of page edits etc etc but you can find those in that archives. Let me know if you need them. --Philip Baird Shearer (talk) 13:33, 3 March 2008 (UTC)

BTW "many" and "some" was a compromise but the editors in the dispute were more evenly spread than that. So one can equally argue it is "some" like CMS and "some" like Nature and the thundering herd don't care. --Philip Baird Shearer (talk) 13:40, 3 March 2008 (UTC)

Thanks for the advice Philip. Yes, if you can provide links to relevant archived sections links, dates of significant page edits etc that will be of great assistance. I realise such stuff is preserved in the archives, but with it being spread over this and the Footnotes article, and with there being so many archives overall, I don't think I'll find it that easy having not been involved with the discussion before.

I suppose what I'd really like to see is a key summary of the main points. I understand such occasional summarisation is recommended practice for lengthy discussion topics. I'd like to understand the reasoning, such as it was. Surely there was something more substantial put forward in terms of argument in a dispute that's gone on for more than a year than it being simply a matter of personal preference.

I don't want to prejudge the issue, but I have some concerns from comments such as SandyGeorgia's above "people just grew tired of the edit warring from a few editors". If indeed, as you suggest, it is not simply the case that a few editors were trying to get their way using brute force with relentless reverts and edit warring until others just grew tired, then it should be possible instead to briefly surmise the actual rationale of the arguments given.

Cheers, --SallyScot (talk) 19:44, 3 March 2008 (UTC)

---

Okay, I've read some of the archives (namely archives #17 & #18) and I can see that it's probably unfair (and unhelpful) to just imply one side in edit warring over the other. However, while I accept that some editors will simply have become exhausted over this and hence understand how the compromise wording came to be, I have to say, for anyone seeking guidance as to appropriate style, it looks rather like a lame fudge. In that respect the "some editors prefer" wording doesn't really resolve the issue, and it's likely to continue getting raised in discussion as people read it and just think, ...huh?

If indeed it turns out we cannot agree a style here then at the very least I think we should somewhere summarise the rationales for the different approaches. That way a new editor has some assessment criteria and can hence judge for themselves as to what might be the best approach for their article. - Which is really what I was getting at with my earlier requests for summary.

One thing that I didn't see enough of in previous discussion was people trying to rationally justify their own POV. Comments such as "Citations/footnotes before punctuation looks awful. This is obvious..." are practically useless in moving things on. If anything such remarks are counterproductive as they only encourage an "oh no they don't", "oh yes they do" pantomime and consequent further entrenchment.

Now, I would suggest that one of the reasons underlying the argument that punctuation before the reference looks bad is because it cuts the punctuation adrift. With the references being superscripted as they are, commas and full stops are inevitably left somewhat floating in space.

It seems to me as if this issue has not been properly recognised or addressed by those arguing for punctuation before the reference. Not only that, regardless of the separate issue of whether Wikipedia house style should allow any approach in this regard, I haven't actually seen any examples which really demonstrate such precedent. The example that's been given over and over is the style of "Nature". However, it seems to me that there is an important difference between typical "Nature" references and typical Wikipedia references; a difference which has a direct bearing on the punctuation "floating in space" concern.

The point is that punctuation will inevitably be cut further adrift by longer reference notations. And Wikipedia reference notations will be at least three times, and typically four times, the length of "Nature's". Wikipedia's inclusion of square brackets practically ensures that by itself. "Nature" references are of course not parenthesised. I don't know why this difference in parenthesis hasn't been picked up already. But in any case, and in addition, I'd also suggest that you'd be hard pressed to find many articles in "Nature" which extend to double digit references, whereas this will be typical for mature, well referenced articles in Wikipedia.

The guideline wording saying - "Some editors prefer the style of journals such as Nature" is trying to suggest that this¹, in effect, is really the same as this^[10], or even this^[12][34].

I doubt if what I've written above will alter positions which have already become entrenched, but, as I say, some summaries of the main points and counterpoints could be included perhaps on a "Further considerations" subpage and referred to by way of a note / link from the main project page. This would inform editors of the reasoning that has been put forward on either side. It would hence facilitate assessment as to the relative merits of each approach, and thus help editors decide overall for themselves what might be best.

--SallyScot (talk) 23:45, 4 March 2008 (UTC)

See:

• Wikipedia talk:Cite sources/archive17#"Footnotes come after punctuation" (June 2007) for what at the time was thought to be the agreed compromise wording, that is roughly what is in the guideline now.

But see also:

The prescription on where to place references tags was added to Wikipedia talk:Footnotes by SlimVirgin at 05:38, 17 May 2006 without any agreed consensus to do so. The prescription on where to place references tags was added to Wikipedia:Citing sources by SlimVirgin at 18:54, 29 October 2006.

• Wikipedia_talk:Footnotes/archive3#citation_location (March 2006)
• Wikipedia_talk:Footnotes/archive3#Citations and punctuation (April 2006)
• Wikipedia_talk:Footnotes/archive4#Period/Full Stop and reference location (May 2006)
• Wikipedia_talk:Footnotes/archive4#Every sentence should have a reference? (May 2006)
• Wikipedia_talk:Footnotes/archive5#Ref after punctuation and the no consensus trap (May 2006)
• Wikipedia talk:Footnotes/archive5#Position of footnotes (July 2006)
• Wikipedia talk:Footnotes/archive6#American/British style (March 2007)
• Wikipedia talk:Footnotes/Archive 7#Footnotes are placed outside punctuation this includes: (June 2007)
• Wikipedia talk:Footnotes/Archive 7#Where to place reference tags (July 2007)
• Wikipedia talk:Footnotes/Archive 7#What do we agree on? (straw poll on ref tag placement) (July 2007)
• Wikipedia talk:Footnotes/Archive 7#The fundamental issue (July 2007)
• Wikipedia talk:Footnotes/Archive 7#Compromise (August 2007)
• Wikipedia talk:Footnotes/Archive 7#Tags placed after punctuation: continuing a previous dispute (8 November 2007)
• Wikipedia talk:Footnotes#Referenced at the end (December 2007)

• Wikipedia talk:Cite sources/archive14#Footnotes come after punctuation (October 2006)
• Wikipedia talk:Cite sources/archive16#"Footnotes come after punctuation" (May 2007)
• Wikipedia talk:Cite sources/archive17#"Footnotes come after punctuation" (June 2007)
• Wikipedia talk:Cite sources/archive18#Placement of footnote reference tags (July 2007)
• Wikipedia talk:Cite sources/archive18#Always place references after punctuation (July 2007)
• Wikipedia talk:Cite_sources/archive18#Refs after punctuation (July 2007)
• Wikipedia talk:Cite_sources/archive19#Reference tags (October 2007)
• Wikipedia_talk:Cite_sources/archive20#Deterioration (January 2008)

Also:

• Wikipedia talk:Footnotes/Archive 7#Reference tags and spaces (19 November 2007)

The reasons for the preferences for different styles are spelled out in detail in these archives. But if you want them listed out yet again I can do so. I would point out to you that the wording of "many editors prefer ...CMS" and "some editors ..." was a compromise as Septentrionalis has pointed out it is actually "some editor prefer ... CMS" and "some editors ..." is a more accurate reflection of the number of editors involved in this debate. --Philip Baird Shearer (talk) 10:57, 7 March 2008 (UTC)

Rather than be dogmatic, it makes sense for the location of a reference relative to punctuation to depend on what element is being refereneced. I.e., if a particular reference supports an entire sentence, the reference should be after the full stop. If a reference is used to support only a particular word or phrase, it should go directly after the element it supports. E.g. "The ball is big,^[1] blue, and heavy^[2].^[3] ... 1. A. Smith, "The ball is big"; 2. B. Jones, "the ball is heavy"; 3. C. Baker, "The big blue ball is heavy""--Jeffro77 (talk) 05:03, 8 March 2008 (UTC)

What do other style guides say?

Thanks for the links Philip. Hopefully I'll get a chance to have a read through more of them. Though I'm not sure I'll manage everything that's been said since the year dot. At this stage I'm a bit concerned that some contributors on this subject have been abandoning good principles of encyclopaedic research in favour of their own POV+OR.

I know there's more latitude in discussion pages, and I realise we're talking about guidelines rather than a regular article here, but really, if more editors thought along the lines of citing their sources instead of simply making stuff up, I'm sure this debate would've moved on quicker. For example, Jeffro77's point above - "it makes sense for the location of a reference relative to punctuation to depend on what element is being referenced" - I have to ask... where does that come from?

With regard to any issues of this sort, the case will always be stronger with good supporting references to existing style guides, ones well-established and already out there which set such precedents. One would still be entitled to argue that sticking to a single house style would be preferable, even if that style choice was somewhat arbitrary, but it would have to be conceded that the case for alternatives would at least be stronger if it could be shown that they weren't simply on-the-spot contributor inventions.

With this in mind I Google searched: footnote number punctuation

I got a number of results, in addition to Chicago Style, including University style guides such as the University of Minnesota and the Australian National University, and other institutions such as the International Development Research Centre (IDRC)'s style guide and the Energy Information Administration (EIA) style guide.

It could be argued that some of these bodies may simply be copying Chicago and in that sense they're simply multiple references to the same style. Perhaps more significantly then, it seems as if other major guides including Dictionary.com, APA and MLA, also suggest that superscripted numbered footnotes are placed after punctuation.

I haven't yet found any major style guides that explicitly recommend superscripted numbered footnotes before punctuation. So it could be suggested that the practice of the journal Nature doing this looks rather idiosyncratic in this context.

Does anyone know if 'Nature' has a formal style guide published as such? Or, in lieu of reading back through further archive discussion on this, do we have any other style guides out there which support the adoption of the 'superscripted numbers before punctuation' approach?

--SallyScot (talk) 14:21, 16 March 2008 (UTC)

Yes there are others (Eg an EU one), there are others (journal of the ICRC) that leave it to the editor to decide, please read the sections I have given you to read to find the links to those and others. Why are you so dogmatic about this? It seems to me that you wish to do away with a compomise that took a long time to agree on. --Philip Baird Shearer (talk) 13:51, 19 March 2008 (UTC)

If Sally really is doing a lot of work and making a good-faith effort to contrast and compare style guidelines, that would be the opposite of "dogmatic", but I don't know the history of this argument. Potentially, the consensus here could be affected by Version 1.0. Please see the discussion below at WT:CITE#STYLE1.0. - Dan Dank55 (talk) 16:08, 19 March 2008 (UTC)

May I suggest you read the links above for a history of the dispute? --Philip Baird Shearer (talk) 14:50, 20 March 2008 (UTC)

---

I'll try and get through as much as I can with the archives. But I invite other contributors to summarise previous arguments in the hope of eliminating some of the prior repetition, static and noise.

Now, as Philip quotes the EU guide above, I can take it that he considers this one of the stronger examples. Reconfirming its candidacy in this way clarifies for me that it's something I should consider seriously and look further into.

I meant this one: "3.4.1. Citations and recitals (preamble)" as well as the one you mention below "8.1. Footnote references" and not the one specifically for bibliographies. Indeed as pointed out in an earlier thread the EU seems to have no problems with different forms of citation for different types of documents.--Philip Baird Shearer (talk) 10:40, 28 March 2008 (UTC)

What you pointed out in an earlier thread (archive 17) was erroneous. There you said that page was "a page specifically about writing Biographies, not any other type of EU document." The guideline is of course not specifically about writing biographies, neither is the EU guideline particularly suggestive of different forms of citation for different types of document. Their style guideline suggests different formats for bibliographic references in comparison to other types of footnotes, including EU 'citations'. The EU's use of the term 'citation' is more in a legal sense. It's not the same as Wikipedia's and is not interchangeable with our encyclopaedic sense of a bibliographic reference. Their 'citations' refer to the EU's own legislative documents, such as Official Journal treaties, accession acts, agreements, protocols, and conventions. Any other type of reference, i.e. to that of the work of another author or organisation (i.e. what we might also call a citation) would fall under its categorisation covered by section 5.3.4 Bibliographies. --SallyScot (talk) 22:33, 1 April 2008 (UTC)

So, from Eurpoa (the portal site of the European Union) - Interinstitutional style guide - 5.3. Preliminary pages and end-matter - 5.3.4. Bibliographies...

There are two different systems for presenting the references in the text for a bibliography; one is numerical (the numbered system), the other is alphabetical (the Harvard system). In the numbered system, references are indicated by numerals in square brackets, placed after the punctuation, and the bibliography is printed in numerical order, also using the numerals in square brackets.

- Note my emphasis on phrasing above. This is quite explicit about the placement of references with respect to punctuation. It doesn't mention the use of superscript and doesn't include an example; it may not look quite the same as Wikipedia, but in any case, it clearly contradicts the 'references before punctuation' argument here.

However, the EU style guide does make a distinction between bibliographic references and other types of footnotes (an issue also in discussion on this talk page further below), saying that one of the footnote forms is...

figure in superscript between parentheses with same value as the text, preceded by a fine space and followed by any punctuation:

References to the Commission Regulation (¹) also appear in the Council communication (²); but not in text of the Court of Justice (³).

Here note that the parenthesis are clearly not superscripted. I suppose this is one way of avoiding the issue of punctuation floating in space, but I'd suggest it isn't really applicable for Wikipedia, given that our square bracketed numbers are wholly superscripted, and there being little likelihood of this changing just to follow an EU style.

A particular reason for bringing up the issue of further examples is that our guideline wording says...

Some editors prefer the style of journals such as Nature, which place references before punctuation.

...and the "such as" phrasing looks rather like weasel words unless some other supporting examples can be found.

Once again please, can anyone provide examples, other than the journal Nature, which put wholly superscripted references before punctuation? - I'm looking for established style guides that explicitly describe such practice. And yes, I have been searching myself, but so far this has only resulted in me finding further style guide examples in support of putting them after punctuation.

Thanks, --SallyScot (talk) 13:19, 23 March 2008 (UTC)

"but I'd suggest it isn't really applicable for Wikipedia" is of course your opinion :-) --Philip Baird Shearer (talk) 10:20, 28 March 2008 (UTC)

I suggest that using parenthesis which are not superscripted like the EU do for footnotes isn't really an applicable way of avoiding the punctuation 'floating in space' issue here because Wikipedia's ref tags look like this.^[99] I'm not really sure what your counterargument would be. Are you suggesting that editors could include them manually and additionally like this (^[99])? Or are you suggesting that some further development work needs to be done to change the way Wikipedia's ref tags work purely to support the EU style? --SallyScot (talk) 17:26, 28 March 2008 (UTC)

I am not suggesting that we use one or the other. I was just pointing out that that is a style guide that places numerical footnotes before punctuation. I am not suggesting that we ape any particular style. See below for my comment about a serious problem with using the CMS on paragraphs that change over time. --Philip Baird Shearer (talk) 09:58, 29 March 2008 (UTC)

I asked for style guide examples that explicitly suggest putting superscripted references before punctuation. My point was clearly about the EU's use of parenthesis which are not superscripted. You suggested the EU guide, so I assumed it to be a good candidate. Yet it advises bibliographic numbered references in square brackets after the punctuation, and only includes footnote reference numbers before punctuation with full (not superscripted) parenthesis. On that basis the issue of superscript does seem to have significance. And, if anything, it looks like the EU has specifically adopted an approach which acknowledges this. --SallyScot (talk) 20:14, 31 March 2008 (UTC)

I also have not found anyone or anything else (a manual of style or otherwise) that follows Nature's example. - Dan Dank55 (talk) 14:00, 23 March 2008 (UTC)

Please see the earlier discussions on this subject I provided an ICRC journal article which uses the same style as Nature. There are other articles other that use other styles as the ICRC journal style guide does not specify where to put the superscript notes. I am sure that other journals could be found, but I see not point when guidelines are meant to suggest styles that editors use, and clearly in Wikipedia more than one style is used. For scientific articles about subjects that clearly ones that appear in Nature, editors and readers are likely to be more familiar with Nature style citations and I do not see why this guideline should force on them a style that they are not familiar.--Philip Baird Shearer (talk) 10:20, 28 March 2008 (UTC)

Following SallyScot's argument, Harvard referencing or non-superscripted before-punctuation referencing should become SOP - they are the most widespread styles by far on a global basis. Superscripted numerical after-punctuation referencing is not very frequently seen outside American English sources, and when it is it is mostly limited to the social sciences, which make up only a fraction of the total material which is citeworthy on Wikipedia. I estimate (from a very quick-n-dirty check in my library) that some 75% of natural science journals use Harvard refs, and the rest is about equally divided between some sort of after-punctuation (almost all US publications) and some sort of before-punctuation (almost all non-US publications) numerical footnoting. I do not think it is wise even to recommend a style that is largely limited to one particular country and only predominant in one particular subject matter. Dysmorodrepanis (talk) 21:41, 28 March 2008 (UTC)

---

Arguing that Harvard referencing or non-superscripted before-punctuation referencing should become SOP is not really following my argument at all. Others have successfully argued for the inclusion of Harvard referencing in Wikipedia on the basis that it is widespread; if you want argue that Wikipedia ought to support non-superscripted numbered referencing on the same basis then go ahead and argue for it. The fact remains that Wikipedia's footnote references are otherwise currently superscripted and it's in that context that the placement of punctuation should be considered.

There are efforts hereabouts to suggest that the use of reference tags after punctuation is something of a Chicago American English oddity "mostly limited to the social sciences, which make up only a fraction of the total material which is citeworthy on Wikipedia". But from my investigations it isn't just the Chicago Manual of Style. Other guides including Dictionary.com, APA, MLA, Oxford/Hart's style and Bluebook legal style also suggest that superscripted numbered footnote references are generally placed after punctuation.

Editors coming to these pages for guidance on such matters might consider what other style guides say in the main to have some bearing. As it stands at the moment I feel that the guidance wording ought to be changed. I intend to include a further note advising of the above guidelines, all of which have explicit advice on the matter, in addition to the existing CMS note. And on that basis I will also reword the advice closer to that prior to Philip's edit of 11:50, 22 March 2008 as I believe the pre-existing was a more accurate reflection of the situation. Also, if anyone wants the "style of journals such as" wording to remain with regard to the journal Nature, then I think it needs to be better supported to avoid the accusation of weasel words.

--SallyScot (talk) 22:55, 7 April 2008 (UTC)

My edit restored a compromise worked out over many months. If you have read the archives, then you will realise that the current wording was a compromise and that no one particularly liked it but was one that everyone could live with. The current wording on the page does not claim that "the use of reference tags after punctuation is something of a Chicago American English oddity" it only claims that some editors prefer CMS and some editors prefer Nature, both of which are easily proved by reading the archives. Please explain how that is using weasel words and please do not change the wording in the section without demonstrating a consensus to do so. --Philip Baird Shearer (talk) 10:36, 13 April 2008 (UTC)

I can hardly stand to see refs before punctuation, but I see no benefit in stirring the issue up again. It led to a nasty series of edit wars over multiple pages, leading to page protection, with no one emerging from it looking very sensible, so I think we should let sleeping punctuation wars lie. SlimVirgin ^talk|edits 10:45, 13 April 2008 (UTC)

---

Philip, I don't think it is true to say that your edit restored a compromise worked out over many months. Your edit of 14:40, 20 March 2008 claimed in its edit summary "The balance for footnote placement. has been tipping in favour of afterendians so put back more balanced wording for beforeedians". This was reverted by SandyGeorgia a few minutes later at 14:43, 20 March 2008 with the edit summary "Restore footnote placement, did I miss the discussion somewhere?". You reverted again claiming "Yes SandyGeorgia you must have done please see the talk page.", suggesting that somewhere in this discussion you believe you've established a new consensus. You changed the wording that said "many editors put the reference tags after punctuation" to say "some editors put the reference tags after punctuation". The many editors wording had in place for months before your change. You seem to have just forced your preferred version through based on your personal feeling that what was already there wasn't quite balanced enough for your liking. You did not get consensus. However, rather than get involved in a discussion about the difference in what exactly was meant by 'some' and what by 'many', I thought I'd do some further investigation as to what a range of other style guides say, in order to contribute something more substantial.

My proposed wording follows from that investigation.

Ref tags and punctuation

Footnote reference numbers may appear mid-sentence, but are usually placed at the end of a sentence or paragraph.

Editors following established style guides, such as Chicago/Turabian, Oxford/Hart's Rules, MLA style, APA style, IEEE style and others, will observe that, except for dashes, superscript footnote reference numbers generally appear after punctuation.

Some editors prefer the approach of the journal Nature, which places superscript reference numbers before punctuation.

If an article has evolved using predominantly one style of ref tag placement, the whole article should conform to that style unless there is a consensus to change it.

Notes

1. ^ The Chicago Manual of Style, 14th ed. 1993, Clause 15.8, p. 494 - "The superior numerals used for note reference numbers in the text should follow any punctuation marks except the dash, which they precede. The numbers should also be placed outside closing parentheses." - See also CMoS Online, Style Q&A, Punctuation.
2. ^ Other established style guides suggesting that superscript note reference numbers should generally be placed after punctuation include: Oxford/Hart's Rules, the MLA Style Manual, APA Style, Dictionary.com, IEEE style and Legal Blue Book Style.

Believe me, if I'd found an even 50-50 split in the guides, I'd be happy to reflect that. I can see why you might want a new editor think it really was six of one and half a dozen of the other, but it's not what I found.

I think if an editor comes here looking for style advice then they may reasonably want to factor this information into their decision. If you think I've specially cherry picked references to some guides and conveniently ignored others and been biased in that way then that would be a different matter. Otherwise I don't really see a reasonable justification in suppressing the information.

--SallyScot (talk) 18:58, 22 April 2008 (UTC)

I would draw you attention to the second from last paragraph in the section #Where to place ref tags immediately above this one that starts "The reasons for the preferences...". If you would like to add a second footnote after the CMS that notes the other styles also use this style,<ref>This is a style also recommended by CMoS Online, Oxford/Hart's Rules, MLA style, APA style, IEEE style and others</ref> I personally would have no objection, but I do object to your changing of the paragraph so fundamentally as it bloats the section unnecessarily, destroys the current symmetry and will encourage others to add many more style guides in an arms race that will not bring clarity. This is after all a style guide and not an article on citation styles. --Philip Baird Shearer (talk) 20:09, 22 April 2008 (UTC)

The part you refer to above "The reasons for the preferences..." may be your justification for making the 'many' to 'some' change, but you didn't necessarily have consensus, and it doesn't necessarily square with your later claim - "My edit restored a compromise worked out over many months". Your objection to my breaking the "current symmetry" seems something of a golden mean fallacy. I didn't want to get into a tiresome debate over what constitutes 'many' and what constitutes 'some', so I've instead proposed a wording which doesn't try to categorise the number of editors preferring references after punctuation. It simply suggests that the 'after' approach is consistent with other style guides. If a style guide 'arms race' ensues and it starts to get silly we can look at it again and deal with it then, for example, by creating a project sub-page for those that want to drill-down to such detail. Who knows, we might even find a widely established guide which explicitly advises superscript references before punctuation. I'd be more than happy to see such inclusion. As it stands though, I don't agree that my edit bloats the section, and I think it is much more informative than the woolly false compromise "some editors" edit.

--SallyScot (talk) 22:48, 22 April 2008 (UTC)

"and I think it is much more informative than the woolly false compromise 'some editors' edit." of course you do otherwise you would not have made the edit! I do not agree with you that it is a false compromise which is why I made the reversal. That you are trying to force through a change suggests to me that you are in support of after punctuation citations and are not interested in a compromise, but wish to force your views on everyone else. It was an insistence on footnotes after punctuation that caused the disagreement in the first place. Why do you think that after punctuation citations are better than before punctuation? Have you read the problems mentioned below that this style causes given the dynamic nature of Wikipedia articles? --Philip Baird Shearer (talk) 11:12, 24 April 2008 (UTC)

My personal preference would be for Wikipedia to have a consistent style. I would comply with references before punctuation if that were the established convention. As I've already said, if I were invited to contribute an article to Nature, then I'd have no problem following their particular in-house approach. However, having now investigated a number of general and widely applicable style guides, it is apparent that the more common convention is to place superscript reference numbers after punctuation.

Even so, though I may argue for it, I have not tried to force through any change to the guideline wording that eliminates the before punctuation option. Your accusation that I'm not interested in compromise is unfounded. My proposed wording does not correspond to "an insistence on footnotes after punctuation". Meanwhile, I have pointed out that your claim above^, that your 14:40, 20 March 2008 edit "restored a compromise worked out over many months", is not in fact true. The compromise wording you refer to actually said "many editors put the reference tags after punctuation" (my emphasis).

Even though I believe it's the case that more editors go with references after punctuation than before, I could see how the definition of "many" could be argued over. I've proposed a new wording which doesn't get into this.

In reverting my initial implementation of new wording you suggested in the edit summary that it had "given undue weight to the after-endians." - A point which remains unsubstantiated. You suggest that including reference to major style guides breaks some kind of preordained symmetry. In this discussion the references after punctuation approach is characterised by yourself as "the CMS style", which you suggest should not simply be aped. But your reaction to other guides being mentioned is now concern about the section becoming too bloated.

Considering the amount of discussion there's been on this topic, brief mention of four major style guides in addition to CMS in the text, with a footnote to support and suggest a couple more, is fairly succinct. It's one single sentence. It's just better substantiated than your "some editors" version (or the previous "many" editors version come to that).

You need to raise a substantial objection to the content as it is proposed rather than change the subject. By which I mean the question - "Why do you think that after punctuation citations are better than before punctuation?" - isn't relevant. Read it again and you'll see that I'm not proposing a wording change which says "after punctuation citations are better than before punctuation". I've simply included some referenced information, worded fairly and neutrally, which adds to the overall value of the guidelines.

--SallyScot (talk) 18:56, 25 April 2008 (UTC)

I did not suggest that you were putting in words that explicitly stated that after-endian is better, but one can make a point through presentation, as any graphic artist will testify. As a compromise solution I have added the additional guidelines. I would prefer not to do so because I think it over eggs the pudding, but I hope you will accept it as a compromise so we can put this dispute to bed. --Philip Baird Shearer (talk) 17:00, 28 April 2008 (UTC)

Thanks Philip. My 'presentation' is only intended to suggest what I found - here that superscript reference numbers appearing after punctuation seems to be the more usual convention. I think it's reasonable for the guideline wording to reflect this. The 'established' wording pre 20th March did this by saying "Footnotes at the end of a sentence or phrase are normally placed immediately after the punctuation…" I'd like to see that restored, with the addition of my new supporting reference.

I realise the process can seem painfully slow, but I have to say I also have issues with the supporting wording around Nature. The style guides suggesting "after" punctuation are more widely applicable than a single journal. As it stands, Nature's approach should really be identified as being a particular in-house style rather than "the style of journals such as..."

--SallyScot (talk) 21:08, 28 April 2008 (UTC)

Citations at the end of a paragraph

My major objection to the CMS style is end of paragraph references. It is impossible to tell if a citation at the end of a paragraph is a reference from the start of the paragraph -- or from the last reference (which ever is closer) -- or if it is just covering the last sentence. This is a particular problem for Wikipedia because at any time an editor may add a sentence to a paragraph which the reference tag at the end of the paragraph does not cover and without a detailed study of the history of the article it is not possible to tell which unless one has access to the referenced material. --Philip Baird Shearer (talk) 10:20, 28 March 2008 (UTC)

Why do you keep repeating this stuff about ambiguous references. It has been pointed out ad nauseum that your proposed solution to this alleged issue would not fix the issue one single bit. Gimmetrow 16:47, 28 March 2008 (UTC)

Eh, it is a problem and it needs to be solved. By whatever means are expedient; unambiguous sourcing is (IMHO) the single most critical point in Wikipedia. Did you notice how the lack-of-quality naysayers have largely disappeared in the last year or so, coincident with the Wikipedia community placing more emphasis on sources?

Every few weeks I come across factual errors, which ultimately turn out to be based on ambiguous citations and sourcing - be it by the Wikipedia editor, be it by the authors of the original sources which have simply not read or ambiguously cited their sources. This phenomenon is considered a pressing problem in the natural sciences; see here for a review. (It may or may not be interesting that they use before-punctuation footnoting despite being UCLA scientists) Dysmorodrepanis (talk) 21:10, 28 March 2008 (UTC)

With "whatever means are necessary" I mean: if Wikipedia - considering the limitations and opportunities given by the underlying codebase which make it precedentless (see also WP:NOTPAPER) - establishes its own style guide, so be it. (From my personal experience with a range of style guides common and in-house, ChiMOS is the last thing I could recommend with good conscience.) Dysmorodrepanis (talk) 21:41, 28 March 2008 (UTC)

Isn't the goal of referencing is to help people who do have access to the referenced material? Anyway, in a paragraph that has a topic sentence, references for that sentence will mostly likely support the overall paragraph, while references on subsidiary sentences are more likely to be for those sentences alone. So if there is a genuine opportunity for confusion, it can be remedied by rewriting the paragraph to make it more clear what the overarching claims are, and attaching the general references to those claims. — Carl (CBM · talk) 20:27, 28 March 2008 (UTC)

I am doing mainly referencing and no source is too obscure or exotic for me as long as it's scholarly. From my experience it would be nice if the approach you suggest worked. However, it does not; on the contrary, simply assuming that the topical sentence is referred by the source at the end of the paragraph is highly misleading. This is because Wikipedia articles are not generally written as scientific articles are (you take a source and reword its content, then you take the next source, etc) but piecemeal; content is added irrespective of relevancy of sources, but rather where it fits best from an aesthetical standpoint. And many paragraphs do not have a topical sentence.

(As I see it, the aim of referencing is to encourage people to read more and educate themselves and do not just believe something just because it's written, without caring about where it comes from. Sources provide a "do you want to know more?" for all sorts of cool and interesting facts. Therefore I always try to provide links to full-text versions of sources.) Dysmorodrepanis (talk) 21:10, 28 March 2008 (UTC)

The goal of citing sources is not primarily about "encourage people to read more" -- that can be done through simple "References" and "Further reading" sections -- it is as a quality check as Dysmorodrepanis pointed out "lack-of-quality naysayers have largely disappeared [since citations started to appear]". On reading a Wikipedia article it would be very tedious to check every citation if all one is interested in is an overview of a subject. I suspect that most readers assume that someone else has checked the validity of the citations in an article. But the dynamics of Wikipedia editing means that although an after punctuation citation may still fully support the sentence next to the citation, one can not guarantee that it covers all the sentences from the start of the paragraph up to the citation. Unfortunately with CMS style citations at the end of paragraphs readers will often assume that it covers the whole paragraph and not just the last sentence (which in many cases, when additional sentences have not been added to a paragraph, may well be true and the intention of the editor who put the citation there). Personally I would like to introduce a hybrid style, to deal with the unusual dynamic environment that is typical of the writing and maintaining of a Wikipedia article, however I recognise that at the moment there is not a consensus for such a combination, but at at least for those articles that are written using the Nature style of citations, there is less chance of misunderstanding than for those articles written using the CMS style, and for those articles written using the Nature style it would be a retrograde step to alter them to the CMS style because it introduces the ambiguity I have mentioned. --Philip Baird Shearer (talk) 09:58, 29 March 2008 (UTC)

Your alleged "hybrid" style does not solve the alleged problem you are claiming exists. This has been pointed out multiple times. Gimmetrow 23:42, 29 March 2008 (UTC)

Does the problem not exist Gimmetrow? --Philip Baird Shearer (talk) 09:42, 31 March 2008 (UTC)

Don't shift the topic. Whether the "problem" exists or is significant is not the critical issue here; your proposed "solution" doesn't fix it. Gimmetrow 17:33, 31 March 2008 (UTC)

I am not shifting anything, I am asking do you think it is a problem because as you wrote "alleged problem" you might not think it to be a problem. Does the problem exist Gimmetrow? --Philip Baird Shearer (talk) 22:28, 31 March 2008 (UTC)

Yes, you are shifting. You point to an alleged problem, and then present a "solution". You have continuously been challenged for months to demonstrate that your solution even fixes the problem you claim exists. Not only have you utterly failed to make any such demonstration, you do not even change your argument to reflect the fact that your solution has been challenged on this precise point for months. Gimmetrow 18:38, 5 April 2008 (UTC)

Oh cut it out already! The question to ask here is "What can be done to alleviate the situation?", no? "How can it be achieved that to the average reader, references are as clear and as unmistakeably placed in regards to referenced claims/statements as by any means possible?"

"The average reader", I'd guess, is likely to be from Europe, North America or Australia but may come from anywhere in the world and is at least capable of understanding Special English.

From the growth pattern, language and style of "amateur" contributions - basic formatting, no fancy tagging, reffing, catting, templateing etc - as far as I can observe them I would say that there is a general growing popularity of the en.wikipedia in (S(E)) Asia. Which may raise the stakes here; one can assume that someone from say Brazil will have no problems understanding any "Western" referencing style, but we might be getting many new members who are far from being embedded in Western culture and might not understand just any style with ease. I don't know if I'm correct; to seek out input from novice users may well be worthwhile. Dysmorodrepanis (talk) 05:49, 1 April 2008 (UTC)

What is your point? It seems to be irrelevant. Gimmetrow 18:38, 5 April 2008 (UTC)

Can you contribute anything or are you just here to troll? Dysmorodrepanis (talk) 14:40, 7 April 2008 (UTC)

You interjected comments into a discussion that appear to be completely unrelated to that discussion. Care to explain your comments, while respecting WP:NPA? Gimmetrow 16:09, 11 April 2008 (UTC)

←Let's ask the question another way: what you're saying seems to be incomplete, Dysmo. Can you point to an article or a series of articles that seem to be a problem? I hate to be self-promotional (no I don't), but I believe Good article usage is about the right level to attack problems like this. We should keep a firm eye on changing trends in layout, wikification and language usage, but the study needs to be large, random, and careful. Take away any one of those adjectives, and what you've got isn't very useful. Feel free to come help out by reviewing articles which have been newly nominated for GA status, which is probably as close to random as we need to get. There are a few more details in the long discussion (and support) at WT:MoS, but WP:GAU reasonably covers it. - Dan Dank55 (talk) 15:45, 7 April 2008 (UTC)

It can not really be random as it takes a specific set of circumstances to throw up the problem. The article needs to carry sufficient citations of the CMS type that it could be taken for granted that the citation at the end of the paragraph refers to the whole paragraph, but that paragraph has had text added that is not covered by the citation. Typically these are controversial pages edited by a number of editors who are not very familiar with citing sources. Articles up for Good Article are usually in fairly good fettle before they are nominated and are not usually controversial or still under development (they tend to be fairly stable). I'll have a look through a couple of candidates that I can think of, but most of it would be historical (as I fix such things when I see them) and I have more pressing ways to spend my time. Next time I come across an example though I will be sure to place it here :-) --Philip Baird Shearer (talk) 11:08, 9 April 2008 (UTC)

Here is an example notice the movement of the Paddy Ashdown sentence from below the citations to above. It is impossible to tell from the arrangement after the edit that the citations at the end of the paragraph do not cover the Ashdown sentence. --Philip Baird Shearer (talk) 17:29, 15 April 2008 (UTC)

Try for example Protoavis, where precise referencing is crucial because this thing is one huge controversy (luckily, not the article - but the sources are full of mud-slinging). I have footnoted some sources; somebody had added a whole bunch of refs that are not cited already, so I suppose eventually it'll be more than just the few cases where it's ambiguous. I guess that the good Dr Witmer would not find it amusing if his name is connected with any of the reconstructions of that thing (as per the Zhou quote). That is a good example - you have a paragraph that ends on a contradictory statement. If properly shourced, the paragraph itself would probably be full of Chatterjee references - so full that one could actually use "The Rise of Birds" as default ref for the entire para. Dysmorodrepanis (talk) 22:03, 21 April 2008 (UTC)

Discursive notes

...earlier part archived...

'My' version looks and behaves more like <ref> tags. Bill's is probably easier to code. But the main point I think we'd both agree is that both (both 'my' HTML and Bill's {{ref label}}+{{note label}} solution referred to in archived part) require the numbering to be manually kept in sync. All I was really trying to suggest is that the ordered list part should be fairly easy for a wiki-developer to do as it would just require style="list-style-type: lower-roman" instead of the default.

To be clear, the request would be for code so that something like the following...

Example text,<nb>This is an example discursive note</nb> more example text.<nb name=Discursive>Discursive notes can be shown separately from references or citations - giving a neater appearing alternative compared to having mixed "Notes and references" or "Notes and citations" sections. This is an example of such a note. It is wishfully generated via a companion to the ref footnotes method (i.e. via use of nb and notes/ tags).</nb> A point made with a supporting reference.<ref>Author, A. (2007). "How to cite references", New York: McGraw-Hill.</ref> A second appearance of a note.<nb name=Discursive /> 

== Notes ==

<notes />

== References ==

<references />

And maybe we could have curve brackets instead of square for some further distinction, producing...

Example text, more example text. A point made with a supporting reference. A second appearance of a note.

Notes

References

--SallyScot (talk) 18:38, 21 April 2008 (UTC)

I think the point with Bill's is not that it's easier to code, but that it already exists. 86.44.30.169 (talk) 02:41, 23 April 2008 (UTC)

I wouldn't be asking for <nb> tags that handled auto numbering if they already existed. --SallyScot (talk) 13:17, 4 May 2008 (UTC)

We shouldn't abbreviate journal names

We often read that something was published in J. Am. Phys. Soc. or something like that. That's standard in scholarly journals. I'd prefer to have a policy against such abbreviations, in favor of writing Journal of the American Physical Society instead. Wikipedia does not have the limitations of print journals, and hence doesn't have the need for such abbreviations, and often there can be uncertainty about the name of the journal, especially when it's not in one's own field. If such a policy were established, the huge volume of editing to conform to it in all the articles we've already got might require a bot to go through and search. Mistakes might happen and might offend some people, so we'd have to take some precautions against that, maybe even notifying people who've edited an article so they can vet the bot's work. Or the Wikipedian's work.

Opinions? Michael Hardy (talk) 02:01, 14 April 2008 (UTC)

Concur, wholeheartedly. Within a given discipline, many journal titles are familiarly known by an abbreviation, but Wikipedia is not restricted to a single discipline. For example, the abbreviation MLA is used for the Music Library Association, the Modern Language Association, and the Medical Library Association, all of which publish journals (not to mention the Museums, Libraries, and Archives Council, and the Mississippi Library Association, neither of which seem to have a journal, but share the abbreviation).—Jerome Kohl (talk) 03:25, 14 April 2008 (UTC)

Whose time is being volunteered to go back through the 2.3 million pages either writing out journal names or vetting a bot? Are you offering? - Dan Dank55 (talk) 03:56, 14 April 2008 (UTC)

All of us, of course. Millions of Wikipedians. Michael Hardy (talk) 21:04, 14 April 2008 (UTC)

A simpler thing is just to add it to the guideline, perhaps as a recommendation, and let people gradually implement it over time after that. By the way, I also think it's better not to abbreviate. — Carl (CBM · talk) 04:33, 14 April 2008 (UTC)

I would make it a guideline; and support Carl in not strictly enforcing it and go for gradual improvement. Abbreviations are not always consistent; so a bot would be difficult. By the way not all scholarly journal use abbreviations, journals published using APA (american psychological association) style have the full journal title. So yes I also support full journal title guideline. Arnoutf (talk) 06:36, 14 April 2008 (UTC)

Oppose. (Changed from neutral.) I'm not okay with "We've changed our minds, let's go back and change the formatting on all our articles, this looks nicer." I'll ask WP:FAC and other people to join the discussion. - Dan (talk) 21:14, 14 April 2008 (UTC) P.S. I shouldn't have put it that way, people might say I'm canvassing. I made a post on WP:FAC, and when I saw quick response, I didn't notify anyone else. The people who have to deal with these things every minute clearly have to be notified of this kind of proposed change. - Dan (talk) 13:11, 15 April 2008 (UTC)

I completely agree that a complete citation is preferable because our product is geared to a general rather than a specialist audience and full citations are clearer and more understandbable for a general audience. However, because many of our editors are specialists used to specialist methods of citation, and because abbreviations are standard practice in the specialist world, using abbreviations which are standard in the field shouldn't be made wrong or unacceptable or be a basis for removal. Because it can be hard just to get people to give citations at all, the perfect shouldn't be the enemy of the good. So I wouldn't phrase it as "we shouldn't abbreviate" or give people yet another no-no, I'd phrase it as full citations are preferable and try to explain to editors why, and try to educate editors to take a look at the point of view of the readers. Best, --Shirahadasha (talk) 21:16, 14 April 2008 (UTC)

Are you saying that we should write the reference to this article as: Pis'ma v Zhurnal Eksperimental' noi i Teoreticheskoi Fiziki 45:1943,1963  :( Count Iblis (talk) 21:35, 14 April 2008 (UTC)

There's nothing stopping (I think!) the use of templates for this purpose: {{j am phys soc}} could expand out to its full name easily , the above example moreso. (I know certain behavior of wikilinks fail in ref templates, but I think templates remain ok...) --MASEM 22:00, 14 April 2008 (UTC)

Strongly opposed to this notion, which comes around periodically. For entries in PubMed (which is a lot of journals), Diberri's tool has been used on Wiki for years to generate citations, with abbreviations, and changing them all would be a poor use of editor time, since the full journal name is in the PMID link, one click away. For a sample of the unnecessary work this idea would generate, look at DNA, and notice the full journal name is always one click away, on the PMID or DOI link. SandyGeorgia (Talk) 22:18, 14 April 2008 (UTC)

Yes, if you have the PMID or DOI link any additional information is redundant. Add to that the fact that if you Google any of these abbreviations (I just tried it with about ten) you get the journal as the top hit. Therefore, as this information serves no vital purpose, and the abbreviations are just as good search terms as the full titles, I don't see requiring full titles in part of a policy as making things any easier for our readers. However, if Diberri's tool or the Scholar Wiki search engine spat out full titles as default, that is what I'd use, but I certainly wouldn't bother filling them in by hand since that would be a complete waste of time. Tim Vickers (talk) 22:37, 14 April 2008 (UTC)

Further, since (I believe) the PubMed database accessed by Diberri stores the journal names as abbreviations, this would invalidate Diberri's tool, and the natural sciences articles folk would have to start generating cite templates by hand. SandyGeorgia (Talk) 22:53, 14 April 2008 (UTC)

It might be possible to produce a tool that retrieved the PubMed abbreviation and referred to the list of standard PubMed abbreviations (link) to expand the title, but it might not be that straightforward. Tim Vickers (talk) 22:55, 14 April 2008 (UTC)

There is no techical reason Diberri's tool cannot output the full journal name as an option (of which the tool has many). The tool reads an XML document such as this (view source to see the structure), which contains:

<Title>Proceedings of the National Academy of Sciences of the United States of America</Title>

<ISOAbbreviation>Proc. Natl. Acad. Sci. U.S.A.</ISOAbbreviation>

So this could be an option that editors can choose, much like they choose the citation format they prefer. And with Tim's list, there is no reason a bot/tool could not be used to help automate conversion -- though I would be opposed to making it anything other than editor choice on an article basis. Colin°^Talk 23:09, 14 April 2008 (UTC)

Strong oppose - Whilst I accept the initial proposition's observation that "Wikipedia does not have the limitations of print journals" there is a "need for such abbreviations" and that is our poor readers who do not have an unlimited ability to scan through full journal names that may easily stretch to nearly the whole width of a line. Any article that gives a few references to "Proceedings of the National Academy of Sciences of the United States of America" is going to look horrible and we are supposed to also care about copyediting issues as editors, and not just the bulk of text we can type in as writers. What I look for first in a reference is the title and whether it has relevance to the question or issue that I seek to understand better. Next is still not the journal name, but the year; frankly a discussion on best chemotherapy for a cancer written in 1980 is of (almost) no relevance now, irrespective of how good the journal was. I can probably make a reasonable guess at say "Arch Otolaryngol Head Neck Surg", "Am J Gastroenterol", "J. Pediatr. Surg" or "J Clin Exp Neuropsychol" and yet have never seen nor read any of these journals. The last thing I need cluttering up articles is drawn out screens of references and nor would anyone who seeks to print out several articles. Have a guess at “Acad Rev Calif Acad Periodontol” – we can have a good idea, but full title of “Academy review of the California Academy of Periodontology, United States Section, ARPA Internationale” is just silly in our attempts to allow people with 800x600 screens to also read pages. We also forget that the full journal name often includes mention of the publishing medical group; hence “Acta Belg Med Phys” is clearly about Belgium Physical Medicine, yet the full name parameter would show “Acta Belgica. Medica physica : organe officiel de la Societe royale belge de medecine physique et de rehabilitation” or “Int J Dev Neurosci” clearly International Journal of Developmental Neurosciences, but take a look at “International journal of developmental neuroscience : the official journal of the International Society for Developmental Neuroscience” and at 135 characters long, cite 10 articles and the journal name alone occupies 4% of a "reasonable" 35Kb sized article.

The next proposal no doubt will be to insist on full names of authors with expanded forenames, indeed lets never have "et al" as Wikipedia is not short of space. There again, let’s not be miserly and consider why we do not advocate freely including the complete abstract paragraphs of PubMed (it is a publication of the US government after all) ... :-)

Clearly this is not how most biomedical journals format their references, and we have come to accept abbreviations throughout the rest of cite journal format with "12 (6): 25-7" being in full: "volume 12, issue 6, pages 25 to 27". Indeed this is part of the learning curve for people to understand references in the real world when they look at hardcopy papers and their reference list sections, nor do we treat our readers as infants wikilinking on section titles to help them understand what "Symptoms", "Diagnosis" or "References" are, so why oblige as a guideline on full journal names ? David Ruben ^Talk 00:13, 15 April 2008 (UTC)

If a journal is notable enough to be a reliable source, it is notable enough to have a page on Wikipedia. So, when using Diberri's tool, click "Link journal" to Wikilink the journal abbreviation. Then, create a stub for the journal if it doesn't exist yet (using Template:Infobox Journal where appropriate), and create the needed redirect. (Perhaps someday we could include the brackets directly in Template:Cite journal.) --Arcadian (talk) 03:00, 15 April 2008 (UTC)

I agree with this, although unfortunately (in my opinion) some deletionists regularly go after academic journals on "notability" grounds: Wikipedia:WikiProject Academic Journals/Deletion contains a history. The inevitable deletion nom tends to dampen the enthusiasm of anyone inclined to chip away at Wikipedia:List of missing journals. --JayHenry (talk) 03:41, 15 April 2008 (UTC)

I don't see the need for a sea of blue WP:OVERLINKing in citations. SandyGeorgia (Talk) 03:45, 15 April 2008 (UTC)

You're opposed to linking the sources in citations? I have to say I find it extremely useful (about a billion times more useful than linking the dates). I'm not suggesting it be required, but am more concerned about whether or not Wikipedia should have articles on journals. --JayHenry (talk) 03:51, 15 April 2008 (UTC)

No, I'm opposed to bots, scripts, templates or whatever methodology automatically linking journals in citations. If you have, for example, The New York Times used 30 times to cite an article, I don't see the need for a sea of blue to link every occurrence. As in all linking, WP:OVERLINKing should be avoided. SandyGeorgia (Talk) 21:18, 15 April 2008 (UTC)

Well, we've got the overlinking now, because Diberri added automatic links to his tool, and it's not optional (at this point). That means if you use the same journal multiple times in one article, they all get automatically linked now by Diberri, so we get not only a sea of blue, but a sea of red as well. SandyGeorgia (Talk) 05:41, 17 April 2008 (UTC)

This is such an incredible statement that I'm going to add you to my list of Wikipedian absurdities. On second thought, you make a good point about bots doing it, but I certainly think that journals do need to be wikified once per page if they appear. If you're trying to evaluate science, it's helpful to know the track record of academic journals and researchers. OptimistBen (talk) 07:08, 15 April 2008 (UTC)

I think it is overlinking. WP os not a source for tracking journal reputation. Our coverage of topics is much too erratic for thatto make any sense at all. It's an encyclopedia of articles of subjects. DGG (talk) 03:56, 16 April 2008 (UTC)

Encyclopedias serve purposes: they are references. Journals, since they output primary knowledge, are almost always notable. So is their science, regardless of whether the general public is aware of it. If a journal is involved in something controversial or has a follow-up on some big issue, that likely belongs on their Wikipedia page. And surely Wikipedia is not too erratic to cover journals; Wikipedia's erraticness makes it better at covering these things. I don't know if I'm for wikifying all journals. I'm not sure if I'm in favor of bots doing it either; I prefer that journals be cited in-text and wikified there. OptimistBen (talk) 00:58, 17 April 2008 (UTC)

Support 100% This is a general encyclopedia, not a specialist one. The argument that would we be required to include subtitles if we didn't abbreviate holds no water. PNAS has a website www.PNAS.org, and it is commonly written on this. Wikipedians, I've come to notice, tend to argue polarized positions: we can't do something one way, because if we did, then everything in the Wiki universe would be forced into that one method. The full title of a journal, in English, less subtitles, is a courtesy to the general reader glancing through a list of references who wants to pick and choose what to read. I research professional journals for a living and come across names of publications, abbreviated, that I can't make heads or tails of, can't find on the Internet, but need as sources (the librarian has a reference for these occasions.) If Wikipedia's goal is a lot of diverse knowledge, journal titles should not be routinely abbreviated because they hinder access for the laymen. Too many editors on Wikipedia argue that additional knowledge is one click away. I disagree with this. An article should be a complete discrete unit of its own, not a collection of clicks. I write for both a technical and a general audience, most scientists do have to communicate with lay audiences. It's no hinderance ot me to include full journal titles. It's the least of the things at Wikipedia keeping experts away. --Blechnic (talk) 05:42, 15 April 2008 (UTC)

Support In general I think that studies should be cited in the text with the full journal name. In the citation all you need is the title, the link to the source if possible, and the doi/PMID. It's best to try to citations to the essentials to make editing easier. If there is no link/doi/PMID (in other words, it's a very old article), then you need to do a full citation. But it's preferable to reference the author, the journal, and the date in-text. In the future we can use these to analyze past errors and ethical concerns; it will be HIGHLY useful. I agree that we could be doing more articles on academic journals and academics, and that they are deleted far too often. OptimistBen (talk) 06:54, 15 April 2008 (UTC)

In other words, you want inconsistently cited articles, one kind of citation if there's a PMID, another if not. Doesn't sound good. What we have now in medical/biology/etc articles are consistently cited articles thanks to Diberri's tool. SandyGeorgia (Talk) 21:22, 15 April 2008 (UTC)

No, there's certainly consistency in citing internet-available articles one way and the others another way. We should keep in mind that citation templates make editing articles more difficult. Brevity in citations is thus valuable, at least for some (myself). At the same time, PMID/doi's allow even readers of the print version to find the articles. OptimistBen (talk) 00:58, 17 April 2008 (UTC)

Comment Linking in with general use: Pubmed holds medical science related articles. Can you expect an engineer to be familiar with that database. In my field everyone know the abbreviation JPSP, PSPB, JCR and JASP; I guess most of you don't (and anyway JCR can mean any of three academic journals). Pubmed uses different abbreviations to confuse the matter. So Yes, I support full titles as that is clearly non-biased, non-arbitrary and least likely to create ambiguous reference. Arnoutf (talk) 09:37, 15 April 2008 (UTC)

• Strong oppose. I don't mind if either full names or recognised abbreviations are used (preferably without the dots, IMV), as long as consistent within a list. It's very easy to work out what they mean, and for the user, there's this and this. TONY (talk) 09:42, 15 April 2008 (UTC)
• Oppose making this a requirement. WP has a history of allowing freedom in the choice of citation format but cares about consistency within an article. I don't believe any move towards requiring a certain format will achieve consensus. FWIW, WP:MEDMOS has long stated that "Some editors prefer to expand the abbreviated journal name." -- indicating that either style is acceptable, but that neither is preferred by WP. David Ruben makes a good case for why full journal names may be impractical, particularly in a long article with lots of references. Looking to the future, there is no reason why Wikipedia 2010 couldn't take a PMID, DOI or ISBN in the wikitext and format a citation according to reader preferences rather than author preferences. However, given the total inactivity wrt fixing date formatting, I'm not holding my breath. Colin°^Talk 10:54, 15 April 2008 (UTC)
• Oppose Wikipedia is open source and written by volunteers. Everyone hates doing their references and endnote is Horrible. I often copy and paste my citations from my work (all abreviated and mostly harvard style). I am against having a standard for references that has to be mannually imposed: There are few editors that try and keep everything togheter and fixing cleanups (way more important), those people are gold and must not be taxed. A unified standard can be discussed when and only when there is a good bot that autonomously fixes all the references using a database index, like pubmed. I for one will not abide by extended references as it will only mean wasting my time. Sorry --Squidonius (talk) 17:20, 15 April 2008 (UTC)

• Support. I support this along the lines laid out above. As for the cumbersome task of formatting citations etc.: in Wikiproject:Mathematics, we are using a database called zeteo (which I wrote), which stores all the information (about journals, authors, and the references, as well). So it gives something like

• Milnor, John Willard (1956), "Construction of universal bundles. I", Annals of Mathematics. Second Series, 63: 272–284, ISSN 0003-486X, MR0077122

(notice not only the fully written name of the journal, but also the issn, which is another valuable information for locating the journal, etc). Jakob.scholbach (talk) 17:50, 15 April 2008 (UTC)

• Support. This is a small step towards making Wikipedia articles on scientific topics more accessible to non-experts, which I see as a good thing, and one that doesn't add significantly to our work as editors. —David Eppstein (talk) 21:27, 15 April 2008 (UTC)

• Support as recommendation, oppose as anything stronger. The full name does look nicer; that's why I did it that way in my own dissertation. But I'm afraid that even strong language in a guideline has the potential to become a distraction and irritant. Let's put it in a guideline but make it clear that it's just a recommendation (even more so than guidelines already are). --Trovatore (talk) 21:27, 15 April 2008 (UTC)

• Oppose as plain unnecessary hassle. Why not leave it alone? Furthermore, I am of the opinion that Wikipedia should move closer to established scholarly practice and not farther away. If this does go through, please make it no more than a recommendation, and please discourage editors going around changing J Clin Psychiatry to The Journal of Clinical Psychiatry. God knows we have enough lame quasi-edit wars as it is. Fvasconcellos (t·c) 22:34, 15 April 2008 (UTC)

• • Comment This seems to assume old-fashioned print journals are the standard to which other media should conform. Maybe that's true today, and certainly a decade ago, or five decades, etc., but how long will it still work that way? Michael Hardy (talk) 00:05, 17 April 2008 (UTC)

• Oppose (per Trovatore). I think this is a good idea, but it is already a guideline (or was, last time I was up to date) to not use abbreviations. I say, keep it a guideline! That way we have the best of both worlds: well-intentioned (but busy) editors can more quickly cut-n-paste references, and the more detail-oriented folks can fill in the abbreviations if they wish. silly rabbit (talk) 23:12, 15 April 2008 (UTC)

• Concur. Keep it a guideline. --Squidonius (talk) 14:01, 16 April 2008 (UTC)

• Oppose It makes refence lists harder to read by adding a lot of redundant info (since the abbreviations are standardised it is easy to check in google if there isn't a doi or pmid), and it makes it harder to read journals if you've had to learn another system for wikipedia, and the other way around. Narayanese (talk) 14:28, 16 April 2008 (UTC)

• Comment in my scientific discipline (psychology) full journal names is the accepted standard, so using abbreviations demands me to learn the abbreviations, which may not be standardised at all. So your argument could be used both for and against full names. Arnoutf (talk) 15:22, 16 April 2008 (UTC)

Then it makes sense to use full names for psychology journal references. Narayanese (talk) 05:35, 17 April 2008 (UTC)

Euhm that's not what I said, in these journals the author is required to provide the full reference, also when a medical journal is cited. I am in favour of: Always being consistent within an article (cf either UK vs US spelling); If possible be consistent in Wiki. So I would accept full ref use in Psychology Wikipedia articles; and standard abbreviations in e.g. medical articles; but not a mix of styles within an article Arnoutf (talk) 08:43, 21 April 2008 (UTC)

Comment Oh, they're standardized? What is the reference for the standardized names of journals? --Blechnic (talk) 02:12, 17 April 2008 (UTC)

Here is one Narayanese (talk) 05:35, 17 April 2008 (UTC)

One? So how many are there? Do they all agree with each other? Is it inclusive? It looks like, from this resource, that if I want to find what a journal abbreviation means, I have to know what the subject matter is (for example, a biology article may have a medical journal, but would I have to search one of the 11 sources for abbreviations for journal titles in medicine? Or one of about the same number for abbreviations of bioscience journals?)[1]

When I can't find what an abbreviation is for, I have a librarian who looks it up for me. I think she uses primarily one resource for this, but I do use some foreign language journals and some fairly obscure ones, and their titles are not available so readily to the layman as people on this board seem to be saying. I, on the other hand, have to know the full title of any journal I reference. If I, as a writer, am using a journal as a reference for a citation, I have to know what it's full name is. Why not go the sensible way, honor the reader (the person the encyclopedia is written for), rather than the writers? Why not ask the person who already has the knowledge right at hand (the person using the reference) to simply provide it? If it is in the citations list, the reader can see its title. If the writer is using it, they know the title. Technical journals often have very relevant titles, and the reader can decide if they want to access it, or a more familiar one. The writer is giving fuller, more complete information every time they include the full name of the journal for the reader to see, right there, not obscured in its abbreviation. If you've never sat and puzzled through the references in a lengthy research article in Science to get the background for something technical you may not appreciate how much work this is, and not just for the layman, but for the expert.

Consider the audience of Wikipedia. Is it the writer? Or the reader? Is it the technical reader or the general reader? Generally the audience at Wikipedia appears to be the educated general reader. This is the person for whom including the full name of a journal is perfect. It gives him or her sufficient additional information to make an educated decision about the quality of a citation and decide whether or not to pursue it further. The abbreviated title is for the specialist reader of that topic alone, not just for the educated technical reader in general, but for the person already educated in that topic. And, are they coming to Wikipedia for their knowledge? No. I don't get any of my information for my research from Wikipedia. I won't be reading or needing your references, as I can't use them for anything. So, you've designed a system for an audience you don't and won't have. --Blechnic (talk) 05:55, 17 April 2008 (UTC)

PS And references are important to properly written encyclopedia articles. The reference is not a throw-away piece of information for a well-researched, verifiable article. Consider how valuable it is. Is it valuable enough that it should be written out in full and accessible, easily, by the audience? Or is it secondary? It's never secondary if you're not writing original research: you must credit those who contributed. The entire article should rest upon the sources who did the original research upon which the Wikipedia article is based. Don't obscure the sources in any way. --Blechnic (talk) 05:59, 17 April 2008 (UTC)

Support I think journal names should be expanded to full form to be as understandable as possible to the general reader. Gary King (talk) 19:11, 19 April 2008 (UTC)

• Comment - whatever the result, there are probably people at Wikipedia:WikiProject Academic Journals who are willing to refine and implement the required edits. I think linking journals from citations to articles about the journals (and sometimes the authors, if they have articles) is an important part of providing the background to the authors and journals we are citing. If a journal has an article, I would say link and either write in full or abbreviate, like we do with qualifications and titles (eg. MD FRCS, FRS). If the journal does not have an article, either create one (if needed) or write the name in full and provide the abbreviation. Citation style is less important than avoiding confusion over what we have cited. Carcharoth (talk) 07:36, 21 April 2008 (UTC)
  • An example of my idiosyncratic citation style is seen at Thomas Snow Beck. As long as the information makes it into the article, someone else will eventually tidy it up. OK, that's lazy rather than idiosyncratic, but I linked to Journal of the Royal Society of Medicine. I agree that there are concerns for overlinking, and maybe some way should be found to identify all the journal links that come from citation templates, and all those that come from elsewhere in an article. Both sets of information are useful. Carcharoth (talk) 07:49, 21 April 2008 (UTC)
• Oppose. Some journal names are absurdly long and would look poor when cited multiple times. Those people who don't use templates and have to type them in by hand would be likely to make mistakes or shorten very long titles by omitting subtitles, which would hinder finding the journal. I would prefer a guideline which suggested using the long form of the abbreviation, preferably without the points, (for example Proc Natl Acad Sci USA), and generating redirects for journal abbreviations to the full title. That would remove the ambiguity over short-form abbreviations such as PNAS, without taking up too much space. The journal title is generally obvious from the long-form abbrevation, and a glossary of what the common abbreviated words expand to could easily be written. Wikilinking the first occurrence seems reasonable, but guidelines should not encourage multiple linking of the same title in a single reference list. Espresso Addict (talk) 13:59, 21 April 2008 (UTC)
• Support Since Wikipedia is intended for a general audience, I think that accessibility should be the primary factor in deciding which (if any) guideline to use. The space issue doesn't seem that compelling, as Wikipedia is not a paper encyclopedia, and the difference between, for example, Proc Natl Acad Sci USA and Proceedings of the National Academy of Sciences of the United States of America isn't really that much (especially in small type in a list of references). Listing the full name of the journal is much more accessible to a general audience, and thus seems to be the best option. Another good option, as has been suggested, is wikilinking each occurrence of the journal title. That way, anyone could easily find out what on earth something like Z Naturforsch B means. Of course, this requires that every cited journal has an article about them, but I (and others) are working on that. ;-) ~ Danelo (talk) 18:45, 30 April 2008 (UTC)
• Support Wikipedia is a non-specialist encyclopedia and should aim to be as accessible as possible. Make this a MOS requirement for FACs if it's not already. I'm not sure a bot could parse all these (and would probably end up only doing some of each reference list, creating inconsistencies, so I don't think it's a good bot task. Mangostar (talk) 21:35, 4 May 2008 (UTC)
• Support Wikipedia users may not be as conversant with standard sources and their "short hand" abbreviations. The use of full names for jounranls and periodicals will allow a easy reference connection. FWiW Bzuk (talk) 14:25, 11 May 2008 (UTC).

Proposed citation style guideline

In my opinion, we need a detailed Wikipedia guideline establishing recommended citation styles a) that are not redundant, b) that are compliant with widely accepted formats, such as the APA style and Harvard referencing, c) that are possibly dependent on the field of study considered, and d) that are flexible enough to include links to external and internal pages. Note: Whether or not we should use templates to achieve this should also be addressed. To implement the proposed standards, we can start by recommending the agreed upon citation styles for new articles and by encouraging uncontroversial changes to old articles. Some time later, we can add the recommendations to the GA or FA criteria.

Opinions?--Phenylalanine (talk) 01:34, 26 April 2008 (UTC)

Re a) I don't understand what you mean with redundant. b) Harvard is one of 3 allowed citing styles (although the footnote type is more frequently used) So I do not completely understand that this would change anything. c) We already have 3 options that cover natural (footnote type) and social science (Harvard type), although this is not linked to type of article. To be honest I disagree as the encyclopedia should provide information to the interested lay-reader. In other words, if you are a psychologists you should not use Wikipedia (but primary sources such as scientific papers) for psychology topics, but you might be interested in robotics articles (so APA style for the Robot article would be my favourite as reader ;-) d) Http: adresses already allow external linking, wikilinking seems less necessary, but ISBN numbes for example already do this.

In brief, I think most of your ideas are already implemented and I do not see what would change. Arnoutf (talk) 14:07, 26 April 2008 (UTC)

No. What we have now is a free for all system in which any Joe can use any citation style in any article. What I am proposing is a system where one widely recognized citation style, for example APA or Harvard, (or possibly a number of non redundant citation styles such as these) is specifically recommended across Wikipedia. By "redundant", I mean equivalent and interchangeable. For example, the template families "Citation" and "cite" can both be used for the same purposes in the same articles, they are redundant. They also do not comply with any widely used citation formats, such as the aforementioned. The result is confusing for newbie editors and alienating for scholars, who have to learn various wiki-specific citation styles and other styles made up by wikieditors. --Phenylalanine (talk) 17:36, 26 April 2008 (UTC)

For example, the template families "Citation" and "cite" can both be used for the same purposes in the same articles, they are redundant. No they can't; citation and cite shouldn't be mixed in the same article, because they use different styles. See WP:CITE#Citation styles. SandyGeorgia (Talk) 17:41, 26 April 2008 (UTC)

No, what I mean is that they are equally appropriate for the same articles, unlike Harvard and APA, which are not redundant in the sense that in some articles one citation style may be more appropriate than the other. "Citation" and "cite" are essentially redundant in that sense. Of course, articles must not mix different styles. I hope this is clearer. --Phenylalanine (talk) 17:50, 26 April 2008 (UTC)

My personal preference would be a single referencing style fo all of Wikipedia; with a single set of supporting templates. I don't mind whether this is APA, Chicago, Harvard, Science or Nature as long as it is consistent and comprehensive. However, the history of Wikipedia has been fairly organic on this file and it seems unlikely any consolidation will be achieved soon. I am ~very sceptical about a topic specific style as this would open up all kinds forks and/or edit wars arguments (consider Neuropsychology - Should that have a medicine, physics or psychology style??). Arnoutf (talk) 18:05, 26 April 2008 (UTC)

I see what you're saying, and consensus will likely be easier to reach if we allow the use of several non redundant citation styles without imposing "topic specific" style restraints and if the use of templates for this purpose remains optional. --Phenylalanine (talk) 18:23, 26 April 2008 (UTC)

Citation templates certainly need to stay optional. I'm not sure there needs to be a single system, but there's pretty close to a single system among all scholars: it's either footnotes a la Chicago style or author-date a la APA. we should endorse a hybrid system of footnotes and author-date referencing for all articles; do what makes sense. Why? Author-date makes it easier to reference lots of different pages and sections from one very good resource (i.e. a book), while footnotes are handy for the things which don't have to be referenced by the page. As far as I can tell, to reference different pages with footnotes requires an entirely new footnote. And actually the Chicago Manual of Style endorses a hybrid system. See CMOS §16.26 and §16.63. OptimistBen | talk - contribs 18:30, 26 April 2008 (UTC)

I'm going on a long wikibreak. When I come back, I'll file a formal comprehensive proposal (not sure whether I should do it here or at the village pump). I'm ok with both Arnoutf's and OptimistBen's recommendations, so if they decide to go ahead with their proposals, that would be great. Cheers! --Phenylalanine (talk) 13:41, 27 April 2008 (UTC)

←I'm sorry you're leaving for now.I've reopened this discussion a few sections below. ---- CharlesGillingham (talk) 01:41, 30 April 2008 (UTC)

My view on this is generally that people are more likely to add citations if they can add them in whatever style they personally favor, without feeling that they aren't doing it "right". I also find that conformity across the whole encyclopedia is rarely of value to either editors or readers, despite the fact that achieving it usually requires many man-hours or work. As such I would be very hesitant to support a proposal like this. Christopher Parham (talk) 01:48, 30 April 2008 (UTC)

I agree with Parham on this. The underlying question, as I see it, is this: should this style guide recommend a single standard way of verifying all of Wikipedia's articles? I can see arguments on both sides:

PRO:

1. If there was a single standard, it would be easier to learn, so new editors would be more likely to use it, so articles would be verified quicker and easier, resulting in a more thoroughly verified Wikipedia, and therefor a more accurate Wikipedia.
2. A single standard would simply look better and be more professional, giving Wikipedia more authority.

CON:

1. A single standard can't be created by simply changing a style guideline. A single standard requires changing thousands of articles. This an enormous task, but could be carried out by a dedicated group of editors. This would also allow the participation of all of Wikipedia's editors. (See #2)
2. Wikipedia's style policies should represent a consensus of all of Wikipedia's best editors, not just the few who happen to be watching this style guideline on a particular week. Editors who disagree with the style guideline will be unwilling to change their articles and will argue that the style guideline should be changed. If a consensus has not emerged amongst the articles as to which method is best, then a consensus is unlikely to emerge for the style guide.
3. This is an example of instruction creep, which creates problems where there are none and can lead to arguments that cause talented editors to leave Wikipedia for good.
4. If one method is truly superior, then it will eventually be adopted by more and more editors. Enforcing a particular method in a style guide prevents this bottom-up, quasi-evolutionary method from finding the best solution. This violates the very spirit of Wikipedia.

I'm persuaded by the second set of arguments. This issue is best decided by the editors of each article. I support any "dedicated group of editors" who actually take the time to standardize large numbers of articles (provided they always take the time to discuss their proposed changes with the editors who watch those pages, of course). But I don't think a style guideline is the right place to attempt to force standardization. ---- CharlesGillingham (talk) 06:19, 30 April 2008 (UTC)

I'm back, and I'd like to clarify a couple of things. What I'm proposing for the implementation of the guideline is that we require new GA, A, FA and FL articles to comply with the new citation style standards. In this regard, we could create an "official" team of editors dedicated to fixing citations in articles nominated for formal reviews as well as in those article that have passed formal reviews. Also, any disputes among individual editors regarding citation styles would be resolved by administrators in favor of the new guideline. In articles not formally reviewed where there would be consensus to keep the old citation styles, these would stay. This way, people would be allowed to add citations to B-class, Start-class Stub-class articles in whatever style they personally favor, unless individual disputes would occur. But in order for those articles to be "vetted" by a formal review process, the citation styles would have to comply with the recommended formats. This way, B-class, Start-class Stub-class articles will not be required to follow the guideline unless disputes arise, in which case, individual editors who prefer the recommended formats will be free to change the citation styles, but no citations will have to be deleted on the grounds that they are not compliant. This sort of lenient approach, will work far better, in my opinion, than a more restrictive approach in achieving the intended result, which is the implementation across Wikipedia of consistent, professional, widely known and easy-to-learn citation style standards.

The enormity of the task that we face in this regard in no way lessens it's merit. In an ideal world, all Wikipedia editors would be watching this page and commenting on guideline proposals. The fact is only a handful do. Does that mean that we shouldn't try to create standards to improve Wikipedia? I don't think so. My contention is that, if we implement this guideline change, we will gain more editors in the long run than we might lose in the process, because of possible disagreements and disputes, which would be minimized with the approach that I'm advocating. --Phenylalanine (talk) 04:15, 1 May 2008 (UTC)

Although I personally believe that compliance with a specific citation style guideline should be a requirement for new GA, A, FA and FL articles, I also think that simply having such a guideline will likely encourage editors to follow it, whether or not it is in fact added to the formal review criteria. --Phenylalanine (talk) 01:23, 6 May 2008 (UTC)

One of the main drawbacks in the citation style that is incorportated in Wikipedia is the reliance on one style (APA) and disregarding the most common referenceing style, the Modern Language Association (MLA) guide which is predominately used in publishing and academia for social sciences' works. FWiW Bzuk (talk) 14:25, 11 May 2008 (UTC).

Adding a "How to write them" section for Harvard or author-date referencing

Hi. I went ahead and did this edit, but User:Hu12 reverted it and says I should make sure nobody has trouble with it first. Two points: 1) the main Harvard referencing has been retitled to Author-date referencing; it's more descriptive, got more hits on Google, and doesn't have uncertain usage (I'd never heard of Harvard referencing before Wikipedia). 2) The footnotes section jumps right into a "How to write them" section. If people are looking at Wikipedia's citation guidelines, it's likely that they already understand what citing sources means. It's best to get them into the mechanics fast -- that increases the likelihood that they actually will cite things. ImperfectlyInformed | talk - contribs 00:04, 27 April 2008 (UTC)

Given that this is a good change and no one has a comment, I'm putting it back it back in. ImperfectlyInformed | {talk - contribs} 06:35, 30 April 2008 (UTC)

I think the {{Harv}} and {{Citation}} templates have several advantages over the cite id html tag. Would you object to changing this section to recommend their use? ---- CharlesGillingham (talk) 03:38, 2 May 2008 (UTC)

Yes. You can add those in as a separate option, but citation templates should not be required: those both look overly complex, and generate more wordiness on the page. If you want to make them more understandable, that'd be nice but the anchor method needs to stay in as an option. I don't like templates like that -- they take up far too much space (especially in the References section). Doing simple anchors makes a lot of sense; it's very easy to understand. Unless I'm missing something, I see no advantages. ImperfectlyInformed | {talk - contribs} 10:28, 2 May 2008 (UTC)

I'm not a professional editor, but here are two advantages I see: (1) I am sometimes mystified by the hidden meanings apparently intended to be implied by the various italicizings, boldings, parenthesizings, groupings, orderings, etc. done in hand-formatted citations. Looking at the wikitext de-mystifies this if it's a templated citation. (2) If WP ever does standardize on a citation style, a lot of hand-editing of individual articles will be needed to bring hand-formatted citations into line; templated citations, OTOH, can be brought into line by adjusting the template. -- Boracay Bill (talk) 23:16, 4 May 2008 (UTC)

Those are fair points, but until we start requiring templates for regular citations, I don't see why we should require them for author-date citations. As far as the referencing of articles goes, you're right that it can be a little mysterious, but generally the first number is the volume, the second the issue (usually in the parentheses), and the pages have a dash in between them. Does that help? It's no harder to remember than those citation templates -- actually much easier. Also, someone could probably write a bot which could format all citations with anchors or bots in a certain way, could they not? Given that author-date citation have a strict format of [Author (date) 'Title' Journal ect.] it doesn't seem like it would be that difficult. Feel free to add the templates in addition. ImperfectlyInformed | {talk - contribs} 23:49, 4 May 2008 (UTC)

Citation templates: How to use them

Following on the discussion above, I've added a "How to use them" subsection under Citation templates. Since that Citation templates section follows both Inline citation styles and Footnotes, I've made the example I used relevant to both. Feel free to improve what I've done. -- Boracay Bill (talk) 02:14, 5 May 2008 (UTC)

punctuation and cites.

In regard to my discussion for a bot request, I've noticed Wikipedia policy does not as of yet strictly conform on this issue. I believe the Chicago Manual of Style is more encyclopedic and readable (and made to be a standard) than the Nature journal, which isn't necessarily an encyclopedic structure and mimicking a use in a popular publication isn't a valid precedent to starting and encyclopedia standard. It doesn't portray itself as a standard, and I think Wikipedia should make a move to place full stops before & against (i.e. no spacing) the citations. Furthermore it violates proper grammar (and therefore the integrity of the article itself as readable) by truncating the punctuation of the sentence structure and leaving what is conveyed unfinished before it can even be questioned by the cite or tag. (Which I think is impulsively jumping ahead and not a cool headed way to smoothly convey an idea into an encyclopedic context, which I feel violates the spirit of NPOV as well). By this reasoning I think there should be a vote to make this. the standard over this. At least in the permanent regards (I can see it as being transitory as in the "citation needed" in-line tags but not where it is a fixture in the article). 67.5.147.10 (talk) 10:36, 3 May 2008 (UTC)

We just discussed this at great length recently, and I'm not sure this is the best time for another drawn-out discussion on the matter. For my part I disagree with your proposal; I don't see a compelling benefit to readers or editors in requiring the use of any particular format across the encyclopedia. Christopher Parham (talk) 01:28, 4 May 2008 (UTC)

I also agree that imposing formulas may be counter-productive, although since many editors are unfamiliar with style guides and exclusively rely on templates, is there a reliable Modern Language Association (MLA) style guide? The MLA style is widely used for referencing social science works which represents the vast majority of Wikipedia writing. As for guides, the MoS already states that articles can be supported with references in two ways: the provision of general references – books or other sources that support a significant amount of the material in the article – and inline citations. FWiW Bzuk (talk) 04:38, 6 May 2008 (UTC).

The main problem with the ChiMoS is that it's very geographically restricted in use at present. You may not be aware this is you are US American, but it has rather little use outside the US, whereas it is really "the" standard in some branches of science there. From what I've seen, more people on a global scale are familiar with Nature-style footnoting than with ChiMoS-style footnoting. The former is globally widespread, the latter is overwhemlingly common in the US and little-used anywhere else.

This is a huge problem; IMHO the ChiMoS is the worst to use as a general guideline. Not because it's so region-specific, but beause it's so radically so. As an evolutionary biologist, I would say that it mandatorily demands one of the most autapomorphic Standard Englishes anywhere in the world. Where would one stop? The same argument that is rehashed ad nausaeam by SallyScot and a few others can be used to argue for Wikipedia to make certain other types of punctuation mandatory, many of which are to any non-US American clearly erroneous - and the citation style of the ChiMoS is a direct consequence of its general way of dealing with punctuation, not something that is derived independently.

Two examples:

• Serial comma
• Verbatim cites. To add a fullstop inside quotation marks if there is none in the original text (as the ChiMoS demands IIRC) is a butchering and garbling of the source, unfit of an encyclopedic work striving for (and arguably still having some problems with) accuracy.

In brief, were the ChiMOS not so ardent in its promotion of what according to the vast majority of the world's English speakers constitutes bad and even wrong English, I would be more equivocal. But I have professionally encountered most of the MoS commonly used in the natural sciences, and the ChiMoS absolutely blows by comparison when it comes to being exact. I would go as far as to say: if there is any commonly-used MoS that must not be allowed at all in an encyclopedic work, it is the Chicago MoS. It allows more than any other commonly-used MoS to deliberately fake citations and butcher quotations. In the restricted and competitive field of the sciences this may be acceptable. In an open and cooperative project like Wikipedia, the ChiMoS is just a vandal tool. Dysmorodrepanis (talk) 13:00, 11 May 2008 (UTC)

Non-in-line citing

What should be done when in-line citing is not appropriate in an article? In The Tales of Beedle the Bard, all the summaries (scroll down) are referenced by a single link, which I just placed in the References section, since I thought in-line citing was not appropriate here. Is this fine? Also, is there a need for separate sections for the 'standard' references and the link; can't they be included in one single section, entitled 'References' or 'References and notes', like in Caesar cipher? diego_pmc (talk) 11:33, 4 May 2008 (UTC)

They can be included in one section as "references" or "references and notes". Plot summaries are a good example of where inline citation is probably not needed and I don't see why it would be in this case. However, a reviewer of the article would probably complain (incidentally to your question) that the reference is not properly formatted. Christopher Parham (talk) 16:09, 4 May 2008 (UTC)

Question posed

Is there a specific requirement that a bibliography only list the full notations of a Harvard-style citation? I could not find a MoS note but this may exist in an FA, GA or other guide. FWiW Bzuk (talk) 05:05, 6 May 2008 (UTC).

Citation templates considered harmful

I have grown to loath the citation template system, which I believe is seriously degrading the source text of the wiki to the point of inedibility. I am not alone; when this discussion opened in the tech mailing list, it went on for days. But nothing came of it. I want to make sure people are aware of this problem, and hopefully get some sort of solution.

To be exact, it's not a single problem, its really a combination of two different issues. The first is that the CITE template is, itself, complex. Careful editing is required to ensure that a single-character problem doesn't render the entire article unreadable. But because there are so many "subparts" to the template, this sort of mis-edit is all too easy to make. The second aspect is that references are supposed to be placed in-line. Since the REF tag requires the CITE to be "inside it". The result is that, if one uses CITE + REF, the article source text can be made almost completely uneditable.

How big is this problem? Well a good example came up when I wanted to make a few minor edits to the excellent Cygnus X-1 article. Please take a moment and go read it over. It's fantastic. Now click Edit and try to find anything at all. It's completely indecipherable gobblygook. This is precisely the sort of technicality that can scare away the very editors we need to encourage on the wiki.

The general consensus last time this issue came up was that it was fixable through a change to the way REF works. If we could put the "body" of the CITE at the bottom of the article, where it belongs IMHO, the entire problem would go away. Not only that, but it would make finding, updating and editing references far easier than it is today. In a general sense, what we want to do is place all of the CITEs together, with whitespace of course, at the bottom of the article somewhere, likely inside the References section. Then we would "wrap" these somehow so they would not appear (a hidden DIV almost works). Finally we would place REF tags in the body text as we do today, except they would all be the "short form" that we have today <ref name=thingy/>. This makes everything much easier to read, easier to edit, and IMHO, easier to understand. There were several suggestions on how this could be accomplished.

I really believe that this is a significant technical problem that needs to be addressed. How do I go about trying to get some traction on this issue?

Maury (talk) 18:58, 9 May 2008 (UTC)

You're right, but the problem seems to be more how the editors of Cygnus X-1 used the cite template (with each term on a separate line) than with the template itself. It's entirely possible, and in my opinion preferable, to list the terms one after another, separated by the | symbol. In that case, the CITE template isn't much more of an obstacle to editing than the standard footnote entry -- and it does have the advantage of forcing stylistic consistency in the notes.

BTW, it took me a while to accept the CITE template; it lacks the flexibility historians like, but for most published works (which are what we cite in Wikipedia) it seems an acceptable solution. --SteveMcCluskey (talk) 19:32, 9 May 2008 (UTC)

I see your claim, and raise you a John Titor. As you can see in this example, the "inline CITE" doesn't really change the situation very much. I'm shooting for perfection here, and I'll happily accept half-way. Maury (talk) 19:39, 9 May 2008 (UTC)

IONO how many people came up with the shortened notes style independently, but I am one of them and I cranked out a lot of such-referenced articles last autumn to make it more familiar. I came to use this style (I found out later it was already being used by many SocSci people, but primarily in combination with Harvard cites and I usually don't do SocSci stuff) to be able to combine a clean referencing format with annotations, i.e. explanatory footnotes in the strict sense. (For me it's basically a Popperian consideration: I don't care how cool a referencing system works under ideal conditions; my aim was a way of referencing the most outlandish sources without breaking, that additionally looks tolerably well and codes cleanly.)

Shortened notes can be spiced up with whatever you like. I needed something that did not break easily, something to handle things like the poetry in Senna obtusifolia or the diversity of footnoteable information in Buteo or things like Passerine#cite_note-13, and I categorically needed alphabetically-sorted reference lists and so on. And there is perhaps one source in 150-200 where I cannot use it straightforwardly.

But I would not recommend to treat it as one of "the" referencing systems. Shortened notes are really more something like a building block or a foundation that can be used in combination with most all referencing systems to make the code neater and articles better-referenced and more informative. The advantage of the shortened notes style is that it can be used (and I can only highly recommend using it) as a baseline. It is clean and flexible, and it is code-wise nonintrusive, does not require templates or scripting except the barest necessity.

It provides a lot of benefits for the scholarly-minded editor: it is easy to see, at one glance, how much of the article is based on up-to-date sources, and what if any parts would benefit from digging up a newer, better source perhaps. References can be neatly sorted; it is very easy to see if one particular source is missing (this can be hard if the entire source is footnoted, as many editors cut corners to avoid overlong footnotes e.g. by just linking to an online fulltext). Dysmorodrepanis (talk) 14:01, 11 May 2008 (UTC)

I have found the template to be seriousöly lacking and do not use it. For one thing, it creates between 10 and 50% code overhead, and this is simply bad programming style. But more importantly, as soon as you e.g. cite a text with a lot of East Asian or Hungarian authors in rudimentary style (only initials, comma delimitation), you are likely to run into trouble. Ditto nobility-based names etc. The templates work well for most Germanic and many Romance authors; they are liable to produce ambiguous or wrong cites for anyone else. I have noticed this when I came across some citations that proved to be untraceable, because the shorthand style in combination with the template messed up the authors' names. Dysmorodrepanis (talk) 13:16, 11 May 2008 (UTC)

---

I also would like to see the changes suggested above to support the "body" of the CITE at the bottom of the article where it belongs. That said, it's probably worth bearing in mind how the Shortened notes method already described in the project page achieves this. I've copied relevant subsection below.

Clearer editing with shortened notes

Because footnotes work by placing the required content inside <ref> tags within the article text they necessarily break up the text to some degree when in edit mode. Article text can become difficult to read and maintain. In this respect well referenced articles can unfortunately suffer disproportionately in comparison to those not so well sourced. In any case the disruptive effect can kept to a minimum by using shortened notes.

See the "Example edits for different methods" page for some comparative examples using shortened notes and full length references in footnotes. These offer representations of edit mode views with examples of how they render to the reader.

--SallyScot (talk) 19:55, 9 May 2008 (UTC)

Actually I like the "Shortened notes with wikilinks" version very much. However I find the internal portion "too complex", I would prefer a "fully shortened" version like you outline above. The downside to that version, which is also on that page, is that you have to hunt around for the unlinked reference. But that said, I think that it might be easier, technically, to introduce these changes because all it really needs is an additional tag in the CITE and REF to refer to each other...

<ref cite=smith pages=99-100/> would be the "placeholder" in the text, and by adding a similar "name=smith" in the CITE would allow the system to be completely automated. I threw in the page numbers because everyone complains about that :-) Maury (talk) 20:33, 9 May 2008 (UTC)

One can actually simulate, although not perfectly, this concept. See this version of the water memory article for instance. I used a "trick", and if you look at the source text I think you'll agree that it's very editable. It seems quite reasonable to suggest that there will be some sort of marker where the numbered note will appear in-line, and using the short refs makes it almost trivial. Maury (talk) 20:05, 9 May 2008 (UTC)

The solution proposed by you here has been suggested before (about 6 months ago). One of the reasons, I recall listed against it, that it may be instable in certain situations/editors. I do agree however that some kind of "reference library" somewhere at the end of the article, only using pointers inside the maintext would be a great development. This, apparently, needs some hardcore code rewriting (which I don;t know anything about). Arnoutf (talk) 21:13, 9 May 2008 (UTC)

That makes two of us! I did look into the code, and in spite of a relatively knowledgeable background in programming in general, I was completely stumped by the syntax in question and decided it was best left to the experts. :-( I am quite heartened to see that I am not entirely alone in believing this needs some attention though! Maury (talk) 23:27, 9 May 2008 (UTC)

Another argument which came up against this "trick" is that all the footnotes produced contain a nonfunctional backlink. In the example article, the final backlink in each footnote is nonfunctional. Placing the hidden div (<div name=cites style="display:none;"> ... </div>) containing the footnote declarations at the beginning of the article instead of at the end provides the additional capability for the editor to control the ordering of the footnotes by declaring them in that hidden div in the their desired appearance order. In that case, however, the nonfunctional backlink in each footnote would be the initial backlink instead of the final one. Bugzilla:12796 is a proposed mediawiki enhancement which would provide this functionality without resorting to the hidden div "trick" and without producing a nonfunctional backlink. Code to implement this is included there. This was submitted to Bugzilla on 2008-01-26, and is tagged "need-review". -- Boracay Bill (talk) 01:58, 10 May 2008 (UTC)

There's a simpler way to do this: use author-date/Harvard referencing. In fact, if you use anchors and then the cite tag at the bottom, you completely avoid these hassles. Now, you can clutter the visible page up some with authors and dates, but I believe that as an electronic encyclopedia we don't have to do that necessarily. We could just do away with the author and date in-text if we wanted and anchor page numbers to the source; if there were no page numbers to reference we could do the author or some sort of title. We're electronic; the author-date isn't necessary in-text. The CMOS basically says: ultimately do what makes the most sense for your reader, and in this case I think we could use an entirely new system, for the benefit of the readers and writers. Also, I like raw citations. The templates are terrible to read, and terribly wordy as well. ImperfectlyInformed | {talk - contribs} 03:37, 10 May 2008 (UTC)

---

I've got nothing against author-date referencing, but it isn't really much simpler than Shortened notes. Shortened notes can be coded as pretty much the same thing, only with the opening bracket of the author-date reference replaced with a <ref> tag, and the closing bracket with a closing </ref>.

An author-date example:

The Sun is pretty big (Miller 2005, p.23),
but the Moon is not so big (Brown 2006, p.46).
The Sun is also quite hot (Miller 2005, p.34).
== References ==
*Brown, R (2006). "Size of the Moon", Scientific American, 51(78).
*Miller, E (2005). "The Sun", Academic Press.

Same example using Shortened notes:

The Sun is pretty big,<ref>Miller 2005, p.23.</ref>
but the Moon is not so big.<ref>Brown 2006, p.46.</ref>
The Sun is also quite hot.<ref>Miller 2005, p.34.</ref>
== Notes ==
{{reflist|2}}
== References ==
*Brown, R (2006). "Size of the Moon", Scientific American, 51(78).
*Miller, E (2005). "The Sun", Academic Press.

--SallyScot (talk) 10:19, 10 May 2008 (UTC)

And author/date doesn't really work well when you're using a reference for a single item. I really hate having to jump from section to section to read it. Maury (talk) 11:50, 10 May 2008 (UTC)

but how well does "when you're using a reference for a single item" relate to WP, where you are not the only editor of the article at issue?

A-D is less repetitive, however. With your shortened footnotes you would have to glance down at the bibliography, while with A-D you would be taken immediately to the source. I'm not sure I'm following when you say "jump from section to section". If you anchor your A-D references, how is it any different from footnotes? ImperfectlyInformed | {talk - contribs} 15:51, 10 May 2008 (UTC)

Yes, author-date in-text references, if wikilinked, do allow the reader to link straight to references rather than two steps as with shortened notes. The disadvantage with parenthetical systems is that they break the flow of the text for the reader more than less intrusive footnotes references. The effect that this has varies from article to article depending on the number of references of course. Author-date references are fine for articles about uncontroversial subjects which may not require so many references. A footnote system (of some sort) comes into its own the more references are used. --SallyScot (talk) 19:17, 10 May 2008 (UTC)

In a controversial article, readers might or might not find the author-date system more intrusive. If the reader is checking the source of each claim, and the reader becomes familiar with the various source authors, it is more convenient to have the source of the claim right there in the text. --Gerry Ashton (talk) 19:29, 10 May 2008 (UTC)

The option to use inline author-date referencing is available. No technical barrier exists which would prevent use of both inline author-date and footnoted author-date references. Without closely re-reading the current version of this project page, my recollection is that it has little or nothing to say on the subject of inline author-date vs. author-date in numbered footnotes vs. both intermixed. The project page probably should provide some guidance in this area. In practice, when I see author-date, my impression is that it's usually either inline or in numbered footnotes — seldom if ever intermixed — and the choice of inline vs. footnoted seems to be made by the creator of the initial ref, and changing from that initial choice seems to be by consensus (either through prior discussion or by discussion of a reverted WP:BOLD change) on an article-by-article basis. I tend to use footnoted author-date myself, as most pages I edit have preexisting footnoted citations. -- Boracay Bill (talk) 23:24, 10 May 2008 (UTC)

Instead of saying inline, perhaps you should differentiate between parenthetical author-date and footnote author-date (I should have changed "Harvard referencing" to "Parenthetical referencing"). I suspect that's what you mean. Structuring references by author and date is a very simple and useful way to structure references in general because the author and the date are two of the most relevant facts, whether the inline is a footnote or not. I think that both parenthetical in-line cites and footnotes should be allowed on any one page. That way you can avoid repetitive footnotes and give people freedom. The Chicago Manual of Style does not appear opposed to using both for references, although I don't know that I've ever seen it professionally. WP, as electronic, should not be constrained by style: it should be constrained by functionality. And, as I said before, we don't necessarily need to use the author and date inline when your citation is anchored; could just use (p. 32) and let people click to see the reference. I know this sounds radical, but it makes sense. Of course, that does reduce the advantage when you're familiar with authors, and it makes it useless in print...but Wikipedia should not be printed anyway. ImperfectlyInformed | {talk - contribs} 23:37, 10 May 2008 (UTC)

At present, there does not seem to be any citation template that uses the Modern Language Association (MLA) style guide. The MLA guide is widely used and is the most commonly used style guide worldwide in the area of the social sciences which represent much of the Wikipedia material. I find that APA guides are often assigned at university as a "simplified style" most often associated with the sciences. What is often difficult to determine in the APA style is the use of multiple editions as the date is tied to the author note. As well, dropping the place of publication also leaves a gap as many publishing houses operate international offices. Due to these limitations, I often "scratch" catalog or reference source using the MLA guide of: Author, title, place of publication, publisher and date format. The APA citiation template also has certain limitations in its format including the use of ISO dating only which often introduces a jarring element in the article as invariably, two date conventions are in play. I would almost (repeat, almost) accept the APA guide as a standard if the ISO dating could be altered to a more readable m-d-y or d-m-y format. I had earlier asked a question as to why no MLA template exists? Again, why not? FWiW Bzuk (talk) 14:25, 11 May 2008 (UTC).

MLA is certainly not the most commonly used style for the social sciences. It is mainly used in the humanities. Since Wikipedia does more scientifically-oriented documentation than humanities documentation, we haven't added the MLA format. However, I think we should change the "author-date referencing" to "parenthetical referencing"; which is a more global and descriptive term. Under that system author-date is the most common method, but we should also mention MLA style. ImperfectlyInformed | {talk - contribs} 22:17, 11 May 2008 (UTC)

Since I write mainly in history-related articles, I would see the need for the use of the MLA system, especially for editors unfamiliar with the conventions of setting up a reference notation or citation, and rely on templates. One of the common issues that I see is that the information provided from templates are rife with errors, mainly due to inputs being made inaccurately. Again, I tend to "scratch" catalog and do not use the templates, but I can see the need to have templates to suit diverse subject areas, especially in the humanities. Wikipedia had primarily scientific-oriented articles? Really? maybe my reading of articles is too limited. FWiW Bzuk (talk) 03:04, 12 May 2008 (UTC).

I would also like to endorse Shortened notes. They easily resolve the problems detailed at the beginning of this thread and are quite easy to use and edit, even for novices. Madman (talk) 19:34, 15 May 2008 (UTC)

A method that doesn't require big changes

I don't foresee using the proposed shortened notes for articles I am typically involved with. It might be suitable for low-editor-activity articles that are stable. One un-described method (in this thread) is what I would describe as "formatted references & text" which separate the text from the references, and actually do make it easier to edit both text and references, and have been used on very actively edited articles. The good result of my technique for reference-heavy articles, is that the editor can locate the end of any reference easily, since the closing </ref> is on the first column of the editing window. This should make it easier to add or edit the text of the article while noticing where the end and start of each reference is. Here is the format, with a two-reference example:

some text at the end of the sentence.NOSPACE<ref>NEWLINE
body of reference material hereSPACE-NEWLINE
</ref><ref>
body of second citation reference materialSPACE-NEWLINE
</ref>
Start of next sentence.

The result looks like this while editing:

some text at the end of the sentence.<ref>
body of reference material here
</ref><ref>
body of second citation reference material
</ref>
Start of next sentence.

---

Recapitulating:

• The intended result is that the editor can easily find:
  - the end of any reference,
  - the beginning and start of each reference when there are several together at the end of a paragraph or sentence,
  - and not least, easily find the start of the sentences,
  - plus generally all sentences are set-off from the references, instead of all being run-on together.
• The run-on aspect of references and text in a reference-intensive article--is avoided. The run-on aspect is challenge for most editors.

This is not a theoretical item. It's in use here: Dismissal_of_U.S._attorneys_controversy. -- Yellowdesk (talk) 04:06, 16 May 2008 (UTC)

Return to the "shortened notes" topic

I agree with User:Madman, User:SallyScot, User:Dysmorodrepanis who lean towards shortened notes. The short footnotes keep the source as clear as I can imagine and they don't clutter up the text for the non-academic reader. When they are also linked (with anchors or with {{Harv}}) there is very little functionality lost. In this system, the original complaint (that templates take up too much space in the source) is moot. ---- CharlesGillingham (talk) 08:18, 17 May 2008 (UTC)

I think we should stick to the current system of letting the primary editors use what works best for them. I personally find the shortened notes system ugly, non-user friendly, and a pain to use. I much prefer using cite templates within a ref tag, as do many others. Why try to force other people to use such a system when its clear that Wikipedia allows and endorses several different methods. Shortened notes seems to primarily be preferred in topics with primarily academic type editors. Let them be used there then, but I see no reason to try to force all other articles to use it. Collectonian (talk) 08:31, 17 May 2008 (UTC)

I think you might have misread me. I think we're just making suggestions here -- exploring some of the possibilities that an editor could choose. I never intended to suggest that we should "force" anyone to do anything. There are articles where shortened notes are appropriate and articles where they are not. Shortened notes solve a few problems (for example, if you are citing several different pages from the same book, it saves you from repeating the citation for each new page reference). They are useless for articles that use a different source for every paragraph (like articles on current events) or articles that use mostly online sources (no page numbers).

More to the point, shortened notes clutter the source a little less, and cluttered source text is the problem that this discussion was trying to solve. ---- CharlesGillingham (talk) 09:50, 17 May 2008 (UTC)

Okay...cause it read to me like this was an attempt to make this the preferred method. Collectonian (talk) 05:33, 18 May 2008 (UTC)

Just to bring up an opinion (that can be a disturbing image, I agree), contrary to Collectie's reversion to the shortened notes, I think they are actually more readable and less obtrusive then full reference citations. Again, only IMHO. LOL Bzuk (talk) 13:25, 18 May 2008 (UTC).

I don't like shortened notes (too repetitive), but I do want less of the references cluttering up the edit window. Why can't we have some sort of way to build a special footnote reference (the "base" reference) without turning it into a footnote, and instead having it display the reference in full? That way we could give it a name and reference it all we want with the given name in the article, with each reference adding an a b c d ect to the main reference. If anyone's not following me, let me know and maybe I could make it more clear. Impin | {talk - contribs} 19:56, 18 May 2008 (UTC)

• You seem to describe the current status quo, with the use of named references, such as <ref name ='some convenient name'/> . Not sure what your point is. -- Yellowdesk (talk) 02:07, 19 May 2008 (UTC)
  • The difference is that in the current system, the "base" reference creates a footnote. What has been repeatedly requested is a way to put all the base references outside the actual text, for instance in the reference section, without creating anything visible at that location. Christopher Parham (talk) 04:54, 19 May 2008 (UTC)
    • ...such as the cite.php extension submittedinjanuary of 2008 as Bugzilla:12796, including code to implement it, still marked "need-review". -- Boracay Bill (talk) 06:04, 19 May 2008 (UTC)

Retrieval dates for online versions of old printed sources, again

I know this has been discussed a couple of times in these Talk archives, but I want to bring it up again. What is the rationale for requiring access/retrieval dates for online versions of past printed materials?

For example, editors are beginning to link book cites to Google Books. Thus, editors are putting "Retrieved on" on their cites, in addition to the usual author, title, publisher, year, ISBN, and page information. It looks very strange to see a book being "retrieved" ... such a link is just a convenience link (problematic too, given the semi-random way Google Books' "limited view" works); the content of the book is unchanging. If the link goes bad, the rest of the cite remains: an unchanging reference to an unchanging book.

Another case are old newspaper and magazine articles. If a cite gives a 1983 New York Times story's publication date, title, and author, and also gives a convenience link to the NYT archive, what is the value of having the retrieval date for this? The content of the story is fixed and unchanging, and is defined by the print/microfilm version. Again, if the archive goes away, the rest of the cite remains, an unchanging reference to an unchanging story. If the archive gets moved, one would re-lookup the online version by the published date/title/author information; knowing the old retrieval date wouldn't tell you anything.

And there is a real cost to having retrieval dates in place everywhere: to us they take up article edit space, to browsers they increase output HTML space, and to readers they clutter up the cite and can be visually confused with publication date. I understand that retrieval dates are necessary for web pages without publication dates, and arguably necessary for dated news stories originally published online (CNN, current NYT, etc.), but I just don't see the rationale for them in the above cases. Wasted Time R (talk) 23:08, 10 May 2008 (UTC)

It's useful to be able to refer to that date in the WayBack Machine at archive.org. In the case of the NYT archive, we can be fairly certain that those will always remain, but other links won't. It's quite possible that some print sources could be basically impossible (or rather expensive/time-consuming) to track down. People will increasingly rid of print archives. However, if you're crunched for time, do what you can. If it's a podunk town newspaper, put the date; if it's the NYT, don't worry about it. That's my take at least. ImperfectlyInformed | {talk - contribs} 23:40, 10 May 2008 (UTC)

The most common cause of newspaper links going bad is that articles get moved behind pay/subscriber walls. Is the WayBack machine able to show the article anyway, or are they enjoined from making free what is otherwise supposed to be charged for? Wasted Time R (talk) 23:58, 10 May 2008 (UTC)

One of the issues with the citation template is that the nomenclature of "retrieved on" is tacked on automatically and now has become part of the architecture of the citationa as judged by the amount of citation templates in place. I agree that the term looks arcane but with its widespread use, it is hard now to incorporate a "found," "accessed" or "located" tag as an alternative. FWiW Bzuk (talk) 14:25, 11 May 2008 (UTC).

To clarify, my issue is not with what word is used here. I don't think books or old newspaper articles should be listed as "found", "accessed", or "located" either. Those printed sources are unchanging over time; it doesn't matter if you "find" a 1976 book in 1988 or 2008, it's the same book. Wasted Time R (talk) 12:04, 12 May 2008 (UTC)

Absolutely agree on that point, sources that are "fixed" in time, do not require a location date. FWiW Bzuk (talk) 19:26, 13 May 2008 (UTC).

The "retrieved date" merely refers to the convenience link to the online version, and may be safely removed on any cite that is not an online link. That's all. (And if the link goes bad, the dead-tree portion of the cite remains valid.) -- Yellowdesk (talk) 00:40, 16 May 2008 (UTC)

But what's the purpose of a retrieval date for an online version that's just mirroring a print original? What usefulness does it have? What does it tell anyone? Wasted Time R (talk) 04:29, 16 May 2008 (UTC)

• On more than a few occasions I have used the retrieval date for munged references to rediscover the orginal edit that created it, and on more ephemeral sources, search for likely new location for the changing location of the convenience link. In some cases a retireval date indicates when the (changing) source was viewed and relied upon, occasionally important, when the source has changed. It's not superflous, but I would consider it optional.
  Who's to say that even a supposedly fixed archival convenience link will stay that way, and what harm comes from using the access date even there, such as in this example:
  "New Hampshire: Nomination of Bainbridge Wadleigh for United States Senator at the Republican Caucus". New York Times. June 14, 1872. p. 1. Retrieved 2008-05-05. {{cite news}}: Cite has empty unknown parameter: |coauthors= (help) -- Yellowdesk (talk) 14:29, 16 May 2008 (UTC)

• The harm is that the "Retrieved on" takes up extra space (a real issue for our longer, heavily-cited BLP articles) and moreoever is visually confusing — the reader sees two dates, instead of the expected one, and has to figure out what each means, which a possible risk of mistaking the retrieval date for the publication date. In the example of this old NYT story, if the link stops working, it's because the NYT moved its archive or changed its for-free policy on this time period or something like that. If you need to find where they moved it to, you'll do a lookup within nytimes.com using the article's title and publication date; when someone last retrieved it won't matter one way or another. And would you really use a retrieval date for a book, that someone happened to look up in Google Books instead of at a physical library? That really seems offbase to me. Wasted Time R (talk) 21:44, 16 May 2008 (UTC)

• Yes, I would, and have. Especially on heavily edited articles. For the reasons I stated further above: an indicator of when the convenience link worked. I do consider it optional. For example, if some link has an old retrieval date, and apparently not findable by search, then I tend toward deleting the convenience link. For more recent dead links, I'm less likely to remove the link--perhaps the publisher/source is in process of revising the link/location. Essential? No. Useful? Yes. The "retrieved on" is in english, and if using a template, the template does indicate through the parameters how to properly use it. Say more about the confusion you've encountered. (I have to remark, there's plenty of other confusion on articles surrounding refs, such as puctuation, quotations, where to place it and so on, and I've done a fair big of cleaning up other's typos and misplacments on that score. Is this that much different?) -- Yellowdesk (talk) 05:23, 18 May 2008 (UTC)

Proposed rewrite of 'What to do when a reference link "goes dead"'

I propose a rewrite of the article section portion, Wikipedia:DEADREF#What_to_do_when_a_reference_link_.22goes_dead.22. Previous discussions on the topic appeared in the following article talk sections.

1. Wikipedia talk:Citing sources/Archive 8#Sources that are defunct external sites
2. Wikipedia talk:Citing sources/Archive 8#Dead link details
3. Wikipedia talk:Citing sources/Archive 9#What to do when a reference link "goes dead"
4. Wikipedia talk:Citing sources/Archive 14#dead links and MLA style
5. Wikipedia talk:Citing sources/Archive 14#blacklisting known expiring web sources
6. Wikipedia talk:Citing sources/Archive 15#Marking links as inactive
7. Wikipedia talk:Citing sources/Archive 18#Intermediary sites, dead links, retrieval/access dates

Start of proposed rewrite, to replace all the section text after the WP:DEADREF shortcut:

---

When a link "goes dead" (see link rot and Wikipedia:Dead external links), it should be repaired or replaced, if possible. In most cases, one of the following approaches will give an acceptable alternative.

• First, check the link to confirm that it is dead. If the link has returned to service but has been labeled as a dead link, simply remove the labeling. See Template:dead link.
• Find a copy of the referenced document. There may be a copy of the referenced document in a web archiving service. If so, update the dead link to point to the copy of the referenced document.
  • For the Internet Archive, go to http://www.webarchive.org/. Also, see Wikipedia:Using the Wayback Machine. Note that there may be a delay of six months before a recent link shows up in the Internet Archive.
  • For WebCite, go to http://www.webcitation.org, and use their search page.
• Find a substitute for the referenced document. Enter key words or phrases or other content from the cited material into the referenced website's search engine, into a similar website's search engine, or into a general search engine such as Google. (A search engine may hold a cached version of the dead link for a short time, which can help find a substitute.) Or, browse the referenced document's website or similar websites. If you find a new document that can serve as a substitute, update the dead link to refer to the new document.
• Deactivate the dead link, and keep the citation information if still appropriate to the article. (This may happen, for example, when an online copy of material that originally appeared in print is no longer online.) In the remaining citation, note that the dead link was found to be inactive on today's date. Even with an inactive link, the citation still records a source that was used, and provides a context for understanding archiving delays or for taking other actions. In order to deactivate the dead link, do one of the following.
  • Turn the dead link into plain text. Remove only enough of the dead link's wikitext or markup language or URI scheme (square brackets, "http://", and so on) so that clicking on the link does not take you to its destination. This will make the link visible to both readers and editors of the article.
  • Turn the dead link into an HTML comment. Place HTML comment markup language around the link. This will make the link disappear when reading the article, but will preserve the link for editors of the article.

If a dead link cannot be repaired or replaced, consider reworking the article section so that it no longer relies on the dead link.

To help prevent dead links, consider citing reference sources using a persistent identifier such as a digital object identifier, if available; or consider archiving the referenced document online when writing the article section, if permitted by copyright. Also, consider avoiding links to web pages that usually disappear after short periods of time, such as at some news sites.

---

End of proposed rewrite. Please comment on this proposal through 2008-05-20 to build consensus. BrainMarble (talk) 19:13, 13 May 2008 (UTC)

I would prefer to keep the language about removing contentious BLP statements supported only by dead links. Do you believe that your rewrite incorporates any material changes to the actual advice, or do you intend it simply as a rewrite for clarity? Christopher Parham (talk) 22:12, 13 May 2008 (UTC)

I wrote the rewrite proposal after working through a series of backlogged dead links using the original version, and after reading the previous discussions. The rewrite is a clearer version of the procedure, to me; but I'd like to find out how it appears to you and other editors in this discussion, before posting it.

Thanks for your note on biographies of living persons. Would the following statement work, inserted as a paragraph between "If a dead link cannot be repaired or replaced..." and "To help prevent dead links..."?

Whether a dead link can or cannot be repaired or replaced, remember that Wikipedia policy (including policy on sources and biographies of living persons) still applies. Consider doing further edits of the citation and cited material, if appropriate, to improve the article. --BrainMarble (talk) 17:22, 14 May 2008 (UTC)

Referencing

Does a Myspace bulletin count as a reliable reference even if it was posted by a reliable source?RaptorX (talk) 19:59, 14 May 2008 (UTC)

If the source can be written out as reliable source without the myspace connection, it would have a better chance of being accepted as reliable. FWiW Bzuk (talk) 20:21, 14 May 2008 (UTC).

Flash Websites

When dealing with a purely Flash driven official website, is it considered acceptable to source it with the main URL? How should one indicate when "page" of the site the information is being sourced from? As an example: http://www.lastexiledvd.com/ is the official English site for the anime series Last Exile and is being used as a source for the English titles and the English DVD and CD releases for the series. It may also be used later for the character descriptions and other plot related stuff. However, the entire site is in Flash, with no HTML alternative. For now I've sourced it using:

"Official Last Exile website" (Flash). Geneon Entertainment. Retrieved 2008-05-14.

I basically just want to be sure this is acceptable, or if there is something more I should add to ensure its being cited properly. Collectonian (talk) 04:20, 15 May 2008 (UTC)

citation patent

User:Wackymacs referred me to the {{citation web}} type of template to use in history of computing hardware. Is there support for the needs of a patent citation, e.g. {{citation patent}}? --Ancheta Wis (talk) 10:35, 16 May 2008 (UTC)

The analog to {{cite web}} is {{cite patent}}. {{ref patent}} looks superior, though. Christopher Parham (talk) 11:27, 16 May 2008 (UTC)

Christopher Parham, thank you. I incorporated your suggestion. Do you have a recommendation for a Nobel lecture citation. It's for the same article. --Ancheta Wis (talk) 14:43, 17 May 2008 (UTC)

Footnote Fonts? Current footnotes destroy layout, presentation and readability

One of the reason contributors are reluctant to lather their articles with inline citations is that the g-d footnotes destroy the paragraph formatting and readability of the article. The footnotes introduce unwanted extra line spacing that makes paragraphs visually run together and appear very unprofessional in presentation, with random and choppy confusing line spacing. The more citations, the worse this gets. Isn't there any way to code the footnotes so that the footnotes don't devastate the layout in this way? In a printed book or journal article, footnotes are of such a small font that they don't disturb the line spacing. Can't Wikipedia be the same? Is there any code to do this now in a ref tag? Rep07 (talk) 03:26, 17 May 2008 (UTC)

Short answer, no. Using the ref tag = using a foot note. I do not understand your statement about it introducing extra line spacing or making paragraphs run together at all. Never seen any instance of that at all. You can maybe resize them for yourself using monobook.css, if its something that particularly bothers you. Collectonian (talk) 03:29, 17 May 2008 (UTC)

Footnotes don't introduce a whole extra line spacing, but they do make the line spacing wider than normal, at least on Internet Explorer. Thus the gap between two lines with footnotes will be wider than the gap between two lines without footnotes, and it can make paragraph spacing harder to distinguish. This was explained as part of WP:Footnotes to be a bug. See 2nd paragraph.[2] There's a fix for logged-in users, but not casual readers. It would be better if it were fixed for everyone. Ty 06:27, 17 May 2008 (UTC)

Wow! Great fix. I've always hated those uneven paragraphs. Is there any kind of a movement to include this as a part of cite.php? ---- CharlesGillingham (talk) 07:27, 17 May 2008 (UTC)