Wikipedia talk:Manual of Style (footnotes)/Archive 8

Question: Footnotes with no way to validate?
Please see Sandra Pupatello. This is the first time I have come across an article whose references (footnotes) have absolutely no external links as a means for editors to verify the reference cited. Theoretically, anyone could simply create a false entry from the any news source, and place it as a reference to support something they wished to add, and nobody would have a way to verify if it were true. I've seen articles that have inline references, and then an addition "sources" section, that lists books or other printed material not available online, but this is the first article I've seen with 100% unverifiable references. I'm quite familiar with WP:CITE, and realize it is not specifically a requirement that footnotes/references have external links, but is there another guideline that covers this type of thing at all? Thank you for answers in advance. Ariel ♥ Gold 08:07, 13 September 2007 (UTC)


 * Printed material is perfectly verifiable (albeit typically requiring a trip to a well-stocked library to actually verify, should one be so inclined). There are many topics for which online coverage is sparse to non-existent, and thus where insisting on its inclusion would be quite unreasonable. Kirill 11:17, 13 September 2007 (UTC)
 * Okay, thanks for confirming that for me! Ariel ♥ Gold 11:21, 13 September 2007 (UTC)


 * If I may add just one thing here: it's not just paper-only resources that can't be readily validated by everyone - there are also paid subscription-only online resources, also a very common occurrence.
 * Your concern about this issue is quite valid, especially because currently there is no conventional way for editors who can verify such a reference to say "User:JRandomUser has taken a look at this reference and it checks out OK". This leaves other editors thinking "surely someone has checked this", and that is a bit dangerous. GregorB 07:47, 14 September 2007 (UTC)
 * Yes, I've run into the subscription-only issue a number of times, but normally, it is possible to find at least some blurb about the article, first sentence, etc., to at least validate the article is at least about the subject in question. I agree there are concerns, but there are also some resources that are simply not available online, and I realize that. It does beg the question of possible invalid references, but unless someone wants to go actually dig up all the sources from the library, well, even then, I guess there's still no real way to "prove" it, it could just end up a "he says/she says" situation in the end. Interesting issue, to say the least. Ariel ♥ Gold 08:52, 14 September 2007 (UTC)
 * Keep in mind that the principle of verifiability is intended to help keep Wikipedia reliable enough that people will want to read it. If Wikipedia were to adopt a policy that only online sources may be used for research, and that books, paper newspapers, and paper academic journals are not good enough to cite, Wikipedia would rightfully be thought of as a piece of crap and most educated people would just laugh whenever they heard "Wikipedia". The reputation of Wikipedia would be utterly destroyed, in which case, it wouldn't matter that the material in it was verifiable, because nobody would be reading it. --Gerry Ashton 12:11, 14 September 2007 (UTC)
 * Oh, please don't misunderstand, I absolutely know, and agree that all these types of sources are valid, my question simply was if an article contained absolutely no other sources, just lists of non-web references, was there any guideline that suggested that web sources should be added. I've myself, added plenty of books, journals, etc., as references. No worries that I'm thinking they shouldn't be here, I agree completely they should! Ariel ♥ Gold 12:24, 14 September 2007 (UTC)
 * Commenting personally as an editor located on Boracay island in the Philippines &mdash; an editor with zero access to any verification facilities which are not available online, I opine that online verification possibilities are very important to editors who do not have a major university library within easy walking distance. -- Boracay Bill 12:39, 14 September 2007 (UTC)
 * Oh I agree that would be nice, this is why I wondered, if somewhere that I'd missed it, was a passage that stated something to the effect of addition of at least a few verifiable web sources is advised. In the case of the article I listed, I mean, this is a living person, who is a political figure, I find it quite strange that there are absolutely no inline citations given that are to web articles. Surely this person has been covered online, I'd think. This was my reason for asking. But yes I agree, it would be extremely helpful to have online references as at least a small portion, if possible, for those subjects where they are plentiful in other areas, but not as plentiful with online. Still, I understand the policy does not require it, so, I guess we can just encourage online references when/if possible. Ariel ♥ Gold 12:44, 14 September 2007 (UTC)
 * SOrry but I disagree. Most relevant scientific journals are availabe online (with backlog going back ever further). But the argument "I have no access to a university library so hobby websites should get preference over thorough scientific publications" is a very bad argument. Verifiability means that if an editor has relevant access the source is relevant. Scientific sources are (IMHO) the most relevant sources around. If an individual editor cannot check because of access, well bad luck. No reason to let in dubious hobby sites Arnoutf 22:34, 14 September 2007 (UTC)
 * Arnoutf, nobody was talking of "hobby sites" or "dubious hobby sites" until you introduced that. What was being talked about was citing sources which are verifiable online.  Let's presume here that we are talking about sources, whether paper or online, which satisfy WP:RS criteria. -- Boracay Bill 23:13, 14 September 2007 (UTC)
 * These look like a perfectly reasonable selection of Canadian papers. They problably are available without subscription; try your local public library. In Canada, and in major libraries elsewhere, they may well be available in microform; other libraries should be urged to subscribe to Lexis-Nexus and similar services. Septentrionalis PMAnderson 03:32, 20 September 2007 (UTC)

A tip for previewing footnotes when editing one section in a long page
This tip recently appeared on the Help desk: I will see if I can add it to in a suitable way. --Teratornis 17:46, 20 September 2007 (UTC)
 * permanent link
 * I added the tip to: . --Teratornis 18:14, 20 September 2007 (UTC)

Reference tags
See these sections for previous discussions on this subject.
 * Wikipedia_talk:Footnotes/archive3 (March 2006)
 * Wikipedia_talk:Footnotes/archive3 (April 2006)
 * Wikipedia_talk:Footnotes/archive4 (May 2006)
 * Wikipedia_talk:Footnotes/archive4 (May 2006)
 * Wikipedia_talk:Footnotes/archive5 (May 2006)
 * Wikipedia talk:Footnotes/archive5 (July 2006)
 * Wikipedia talk:Footnotes/archive6 (March 2007)
 * Wikipedia_talk:Footnotes/Archive 7 (June 2007)
 * Wikipedia_talk:Footnotes/Archive 7 (July 2007)
 * Wikipedia_talk:Footnotes/Archive 7 (July 2007)


 * Wikipedia talk:Cite sources/archive14 (October 2006)
 * Wikipedia talk:Cite sources/archive16 (May 2007)
 * Wikipedia talk:Cite sources/archive17 (June 2007)
 * Wikipedia talk:Cite sources/archive18 (July 2007)
 * Wikipedia talk:Cite sources/archive18(July 2007)

From my talk page:
 * What compromise wording? It's just you enforcing an editing style that we don't use on Wikipedia.  That makes no sense. We use cite.php referencing after punctuation.  It's really that simple.  And why would anyone use cite.php before punctuation?  It looks terrible on the page, breaks up the flow of the paragraph, and serves no useful purpose. &mdash;Viriditas | Talk 23:01, 20 September 2007 (UTC)

see Wikipedia talk:Cite sources/archive17 for agreed compromise wording.

There are probably thousands of articles where the reference tag is placed before the punctuation. See the Wikipedia talk:Footnotes/archive4 back in May 2006 there were 12,000 articles with reference tags before punctuation. There is no reason to think that there is less now. As I said in this section Wikipedia talk:Cite_sources/archive17 if "you read in a previous discussion that shortly after the prescription was edited in by SlimVirgin that there were about 12,000 articles where the footnote came before the punctuation mark. Even if one assumes that 90% of those were unintentional that still left 1,200 articles that had been deliberately structured that way. Those footnote tags on those pages were probably added by a lot more people than have expressed agreement with the prohibition on placing reference tags before punctuation marks. --Philip Baird Shearer 17:39, 27 June 2007 (UTC)" --Philip Baird Shearer 23:33, 20 September 2007 (UTC)
 * How many articles use reference tags before punctuation at this moment? And how many of those are good or featured articles?  I'm going to guess an insignificant percentage for the former and zero for the latter.  How does this justify changing this guideline?  From what I can tell from the discussion, using cite.php after punctuation is accepted house style on Wikipedia because it works and is embraced by the majority of Wikipedians.  Philip, you've made ~28,309 edits and I've made ~42,640, so we've both seen plenty of different styles.  Can you honestly tell me that there is a need to use cite.php before punctuation?  If so, what is this need and how does it benefit Wikipedia?  Show me the articles. &mdash;Viriditas | Talk 23:45, 20 September 2007 (UTC)

Please read the history of this dispute, there has never been consensus on this issue. I don't think it is "embraced by the majority of Wikipedians" but even if it were there true, there is a substantial minority who use other styles so there is not a consensus on the issue. The wording worked out earlier this year was a compromise. The compromise is that the current style can stay as the recommended one but other styles are not prohibited. And yes I think that more than one style should be allowed for the same reason as that both Commonwealth and American spelling are tolerated with the proviso that articles are internally consistent. --Philip Baird Shearer 00:03, 21 September 2007 (UTC)

The trouble with finding articles for you to view is that articles I remember I tend to have edited and presumably you want to see articles which I have not been involved with. Besides for most of the last year I have been following the guideline in the assumption that I should until it is changed :-) --Philip Baird Shearer 00:16, 21 September 2007 (UTC)


 * If you assume that 12,000 articles use references before punctuation (and assume none of which are GA or FA) and there are currently ~2,016,704 articles, are you actually telling me this means the majority of Wikipedians do not use references after punctuation? Am I reading you correctly?  Perhaps you can explain this to me. (I'm obviously missing the number of unreferenced articles.) Multiple styles should be allowed for different types of referencing formats, but according to Help:Footnotes,  the cite.php syntax is currently the "best-practice method in most circumstances", and that is what we are talking about. &mdash;Viriditas | Talk 00:09, 21 September 2007 (UTC)

I would guess that the majority have no citations at all :-( The thing about GA and FA is that the people who are into that sort of thing are bound to follow this guideline, so if the wording is after the punctuation it will be changed during the process of bringing it up to GA standard. However do you really believe that if scientific journals like Nature consistently use punctuation before punctuation that none of the scientists who contribute to this encyclopaedia and come from such a background would not prefer and place in their articles the form of citation ref tags that they are familiar with? --Philip Baird Shearer 00:16, 21 September 2007 (UTC)
 * I have just come across this article Names of Burma/Myanmar. It only had two reference tags but both of them were before punctuation. --Philip Baird Shearer 08:55, 29 September 2007 (UTC)


 * I'm not sure if this is accurate, but I just read that there are ~79,228 articles tagged as unreferenced. That still doesn't give us a total number, nor is the use of the unref tag entirely accurate in any case.  We really need better numbers to know exactly what we are dealing with.  Considering the talent on Wikipedia, I'm surprised we don't have these numbers at our disposal. (What are the bots doing, if not that?)  Philip, can you address this question: does using superscripted footnotes before punctuation on Wikipedia improve the layout for the reader? What's the point of writing an encyclopedia that is difficult to read?  Layout should be as easy on the eyes as possible, allowing the gaze to flow from one paragraph to the next without interruption.  Now, indulge me a bit:


 * 1. Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua . Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat . Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur . Excepteur sint occaecat cupidatat non proident , sunt in culpa qui officia deserunt mollit anim id est laborum.
 * 2. Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.


 * Which of these paragraphs are easier to read? &mdash;Viriditas | Talk 02:26, 21 September 2007 (UTC)
 * They would appear to be equally easy to read. Though I'm sure some people would prefer one way and some the other. Christopher Parham (talk) 03:45, 21 September 2007 (UTC)
 * "Easy to read" is the goal of the copywriter. In the first example, the punctuation is broken up by inserting superscript in between; in the second example the punctuation follows the text, making the transition smoother for the reader.  Good copywriting places superscript note numbers after punctuation for this reason. This has little to do with using references. &mdash;Viriditas | Talk 03:55, 21 September 2007 (UTC)
 * The basis of the dispute is that different style guides say different things; if there were indeed a uniform standard of good copyediting then the question would not be at issue. And since you already appear to have concluded which paragraph was easier to read, it's not clear why you raised the question in the first place -- but having looked at them both there doesn't seem to be a material difference. Christopher Parham (talk) 04:27, 21 September 2007 (UTC)
 * Brief overview: The Chicago Manual of Style note-bibliography style uses a superscript number after punctuation. Most book publishers use this style, including the Oxford University Press. The other style in question is the citation-sequence system as described in the CSE Manual for Authors, Editors, and Publishers (formerly CBE Manual)  This system uses citations that "immediately follow the title, word, or phrase to which it is directly relevant" rather than at the end of a clause or sentence.  This seems to be the primary advantage of using citations before punctuation, although more needs to be said about this topic.  (If this style can actually improve the encyclopedia as others have suggested, then I will support it above and beyond design and aesthetics.)  This style is often used in scientific research papers and journals.  From what I can tell, the CSE citation-sequence system is flexible enough to allow the use of superscript after punctuation; in some cases the manual explicitly allows for tailoring the use of footnotes for a "house style" although I may be reading too much into this; at least two reliable web-based style guides use the CSE citation-sequence superscript system after punctuation, although there is no clear indication as to why or how:  It should also be noted that the standardization efforts of the CSE have been difficult.  According to Scientific Papers and Presentations (2005), "no consistent standards for style exist in the scientific community" in spite of the efforts of the CSE.  Authors are advised to follow the style guide of the individual science publisher.  Interestingly, the CSE states that in some areas, it is based on the Chicago Manual of Style.  As for a uniform standard of good design practices, Type & Typography (2002) recommends that "the superscript number follows any punctuation."  And, The Art of Scientific Writing: From Student Reports to Professional (2004) advises that "Superscript citation numbers should be set after any punctuation marks present."  The Handbook of Technical Writing (2003) also recommends using the CMS style: "Place superscript numbers at the end of the quotation or sentence after the punctuation marks."  Unfortunately, no standard emerges, as the Mayfield Handbook of Technical & Scientific Writing contradicts the former.  I'm getting the impression this is a battle between North American and European style guidelines, with the latter believing that the use of superscript citations before punctuation is tantamount to  placing punctuation marks outside quotation marks or inverted commas.  There is also the issue of precision. Mig77's proposal addresses this issue nicely, but I'm afraid it would be difficult to implement. &mdash;Viriditas | Talk 06:53, 21 September 2007 (UTC)

What, this debate still rages on? That a "substantial minority" use other styles does not mean much. A substantial minority uses punctuation inside quotes, and I would suspect a substantial minority are familiar with and consciously prefers this. Still, the Wikipedia house style says this is not done. In my opinion, the July 2007 discussion did not represent a consensus for the text PBS calls the "agreed compromise wording". Gimmetrow 00:35, 21 September 2007 (UTC)
 * The evenly divided discussion above seems clear: some editors think it is a house style, others don't. I see no reason we can't just say so. Septentrionalis PMAnderson 16:31, 21 September 2007 (UTC)
 * I see 9-5 for refs after punctuation, but that's another issue. The text you call "agreed compromise wording" was not really at the point of agreement among those involved in July 2007. Then someone started adding it into WP:CITE. Since that generated an edit war, it pretty much demonstrated the text was unacceptable to many. Gimmetrow 23:40, 21 September 2007 (UTC)
 * Oh come on, not again this debate. As long as people use "beauty" as an argument nobody should take this debate seriously. Please close and archive this, as it is going nowhere. (BTW I think the only solution to the re-appearance of this trivial debate would be to allow both before and after punctuation consistently used in the article decided by the first editor implementing footnotes) Arnoutf 10:48, 23 September 2007 (UTC)

What Mig77 said in is what I and other have said before [ ]. However as that is not a style that seems to be followed the Nature style at least implements the precision part and when I raised this issue before it has been suggested that if that is needed then it can be done through multiple citations in any sentence where it is needed. The nature style has the advantage of "fail safe" with regards to this issue.

Viriditas I want to pick up on two things you mentioned above: There is also the issue of precision. and It looks terrible on the page, breaks up the flow of the paragraph, and serves no useful purpose. See above . It does serve a useful purpose which is specific to Wickies.

What Mig77 said in is what I and other have said before (see for example Wikipedia talk:Footnotes/archive4). However as that is not a style that seems to have any published support, following the Nature style at least implements the precision part. When I raised this during the last round, it has been suggested that if that is needed using the Nature style then it can be done through multiple citations in any sentence where it is relevant. At least the Nature style has the advantage of "fail safe" with regards to this issue.

If a person writes a sentence like this:
 * He was six foot six tall and he weighed 225 pounds, but that giant of a man was smitten by love.[1]

or I could write
 * He was six foot six tall and he weighed 225 pounds, but that giant of a man was smitten by love[2].

Now if someone else adds a sentence before it:
 * Smith lived in Canada in the 1950s. He was six foot six tall and he weighed 225 pounds, but that giant of a man was smitten by love.[1]
 * Smith lived in Canada in the 1950s. He was six foot six tall and he weighed 225 pounds, but that giant of a man was smitten by love[2].

How does the reader know if "[1]" refers just to the last sentence or to all the sentences in the paragraph? In the case of "[2]" it is clear that it still refers to the specific sentence, so even if the Mig77 proposal is not implemented it has the fail safe I mentioned above.

It is primarily for this reason that I think the compromise wording is superior to the "house style" wording, although I think that the "Mig77 proposal" is superior to either but as yet does not have enough support to be in the guidelines. --Philip Baird Shearer 20:47, 24 September 2007 (UTC)
 * Philip, when you speak of "Nature style", do you really mean CSE citation-sequence system? Does this include Mig's system?  Since these style guidelines allow flexibility for the house style of individual publications, our focus should be on developing Wikipedia house style that serves our needs rather than copying the style of selected publications. CMoS style is the primary footnote convention for published books (and Wikipedia), whereas the CSE system is used by selected scientific journals like Nature.  If the CSE has benefits that will improve the precision of Wikipedia articles, then we should consider how we can apply them.  But we need to keep in mind what works best for our readers and editors.  I fully support free choice and flexibility, and while the CMoS style  may be imprecise and redundant, it's easy for people to use and it works.  It's also enforced from the top-down.  It occurs to me that we can have our cake and eat it too: extend the highlighting scheme with options so that whenever one clicks on a reference, not only is the publication shaded (current practice) in the notes and ref section, but the relevant text in the body is as well.  This would have the added benefit of encouraging research, closer attention to detail, and allowing the reader to quickly verify the specific portions of the text.  We can also use markup to avoid adding redundant citations, so that highlighted text appears apart from the main. There should also be an option to "hide/show" all citations (particularly by section), so that articles can be read free of superscript.  I apologize for drifting into a tangential cloud of possibility, but we need to keep things as simple as possible.  Philip and Mig77 raise very important points that need to be addressed.  But we should also consider that we are dealing with old media conventions that the new media has the potential to improve upon and extend for our purposes.  If we are going to use a system other than the CMoS, which is simple and easy for the average editor to apply to any article, we should consider where we are.  This is not a print publication.  We have the ability to augment the old media citation standards into a new form that works better than before and I think we should try.  I'll get off my cloud now. &mdash;Viriditas | Talk 22:00, 24 September 2007 (UTC)

I do not know if the Nature style uses the CSE citation-sequence system or places reference tags before punctuation (and not "immediately follow the title, word, or phrase to which it is directly relevant"). Here is a PDF copy of a Nature article (mentioned in an earlier section on this page). There is also the European guideline (also mentioned further up this page) that explicitly says "... and followed by any punctuation".

But nice as your suggestions are for highlighting the text that the citation supports, it is not implemented at the moment, so the change of wording that myself and others wish to include in this guideline, is a compromise which addresses some of these issues. If in the longer term a Wikipedia house style evolves from this then that is fine by me, but the current proposed compromise wording allows for two styles both of which are in current use by third party publications with some extra wording to cut down on pointless edit wars. I do not see why anyone should object to this as the original prohibition on the Nature style was placed into the guidelines without any consensus to do so. --Philip Baird Shearer 14:08, 25 September 2007 (UTC)


 * PBS, however it may have found its way into the guidelines, it was there for ages. You should probably not try to use the "European guideline" as evidence for your case. That guideline says to use "figure in superscript between parentheses with same value as the text ... followed by any punctuation". Superscript between unsuperscripted is not only extremely rare on Wikipedia (and so an example that WP does not use every available style), but the guideline specifies using the ref mark before puncuation in all cases, so it's not at all what you are advocating. Finally, above you claim that in this text:
 * Smith lived in Canada in the 1950s. He was six foot six tall and he weighed 225 pounds, but that giant of a man was smitten by love[2].
 * 'In the case of "[2]" it is clear that it still refers to the specific sentence.' I disagree and have stated so before. From the text alone, I would not be able to determine whether the [2] referred to the last phrase, the last sentence or both sentences. Gimmetrow 13:02, 29 September 2007 (UTC)

Of course you can tell it does not refer to the previous sentence if it did then the reference tag would be included in that sentence as well. --Philip Baird Shearer 20:12, 29 September 2007 (UTC)


 * "Of course you can tell"? Really? Please explain how you know from the text alone. Gimmetrow 01:57, 30 September 2007 (UTC)

Gimmetrow the only reason the other wording was not reverted sooner is because myself and others refrained from a prolonged edit war over the issue. Now what would you consider to be a compromise wording that does not force your views (that reference tags should follow the CMS and no other) on other editors, when it is clear that there is not ca consensus for this. What would you suggest as a compromise that we can build a consensus around? --Philip Baird Shearer 20:18, 29 September 2007 (UTC)


 * I already tried working out a compromise months ago; it didn't work. Given your response above, I don't see much to build on now. Gimmetrow 01:57, 30 September 2007 (UTC)


 * For what it's worth, in my opinion putting the reference link after punctuation looks much better and works well with the general principle here that references should be at the end of a sentence or clause. Putting the link before the punctuation doesn't convey to me that it only applies to that sentence, and the implication is that a reference serving a whole paragraph would have to have a link before every dot and comma. Unsightly and unnecessary. .. dave souza, talk 09:13, 30 September 2007 (UTC)


 * Personally I would prefer inside the punctuation point for a specific reference and after the punctuation for all the information from the start of paragraph or the last reference tag (which ever is closer). But that is not currently on the table as the two styles mentioned in the compromise do not use that style. What is being suggested in the compromise is that the CMS is preferred style, but it does not prescribe the other widely used method as used by Nature, and leaves the style consistent on a page by depreciating changing from one style to the other unless there is a consensus to do so. --Philip Baird Shearer 10:32, 30 September 2007 (UTC)

Citations/URLs
There is nothing in the footnotes section that says that you can not use citations. There is also a glaring omission which is there is no guideline on how to write a URL footnote.


 * External links using URLs should include the URL in full followed by a space and followed by the title of the article in the URL or a portion thereof, inside single brackets, optionally followed by the date accessed:.

If you wish you can add something about using a citation template, but that is already covered, in my opinion, by the statement that references "may be formatted either by hand or with the assistance of templates." If you look through Wikipedia articles it is glaringly obvious that tons of editors have no clue how to format URL references. Please make corrections to the above and put it back. Thanks. 199.125.109.88 01:06, 28 September 2007 (UTC)


 * This is not a footnote but a citing issue: see Citing_sources and Embedded citations. Arnoutf 14:10, 29 September 2007 (UTC)

Can we please discourage the use of named references?
I refer here to the practice of amalgamating identical footnotes, and using a name to send the reader to the same footnote from multiple locations, as described here.

The problem, as I see it, is it's not reader-friendly. The reader clicks on the footnote number to visit the footnote, then tries to get back to where (s)he was, in order to continue reading. But instead of a single, unambiguous "go back" caret, the reader gets an unwanted choice (a, b, c, ...). It would be unreliable, and probably pretty annoying, to use this system to try to recover one's place.

In order to serve our readers well (priority #1, right?), we should simply repeat identical footnotes, so clicking always gets you back to where you were before. The cost in terms of article length will usually be minor. Opus33 18:34, 11 October 2007 (UTC)


 * This was discussed before a little higher up on this talk page, I think it would be better to continue there rather then restarting from scratch. Arnoutf 18:38, 11 October 2007 (UTC)


 * Hello, with all respect I think this is a different topic. The discussion above concerns whether named references make it harder for editors to edit, but I am trying to argue against them because they make it harder for our readers to read.  No one has brought this up before.


 * I feel strongly that we should look out for our readers' interests--that they are more important that we are--and I'd like to know if others agree with me. That's why I would prefer not to bury my comment in the middle of the page.  Sincerely, Opus33 19:27, 11 October 2007 (UTC)
 * I agree, the focus is different; although some arguments maybe similar.
 * On the other hand as reader, I sometimes like to find all uses of one source back easily, the a,b,c's etc. help me there; the same source many times repeated does not. Also if I want to read more, as a reader the most authoritive work in an article is more easily found by looking at the most frequently sourced one. Also as a reader the repeated number shows a repeated source, so no need to check; you have checked before after all. So repeated numbers give allow the reader not to read the references. Finally as a reader I rarely look at the references (unless I am in editing - doubt everything- mode), but even then when reading I use the "back" button on my browser, rather than the numbers and letters; so I think the problem is not very big to start with.
 * So as a reader (and this is of course very tentative because we should realise editors are not good examples of naive readers) I am at most ambivalent, there are several advantages; and the disadvantage does not look too bad. In other words, I think (ok it's my personal opinion, so feel free to disagree) we need not discourage, even stronger I would say we should encourage the use of the named ref, both for editors and for reader convenience. Arnoutf 19:46, 11 October 2007 (UTC)
 * I'll repeat here something I've said elsewhere. Please excuse my beating this dead horse yet another time. I'd like to see an enhancement allowing editors to insert an icon which, if clicked, would attempt to take the browser one page back in its history &mdash; that is, the same action produced by clicking the Back button; but done by clicking an icon in the page rather than going outside of the page to click the Back button. This is easily done with javascript for javascript-enabled browsers (the overwhelming majority) but, as I understand it, use of javascript from wikitext is disabled (and I agree with that). implementing this would not require enabling access to javascript in general from wikitext but would, I think, require either (a) a small change to the mediawiki code done here at wikipedia or (b) that same small change done at mediawiki.  This is a small user-friendliness issue, and I think its user-friendliness value outweighs the fact that it is a small issue and the fact that its implementation (if I understand things correctly) requires a low-level change. -- Boracay Bill 23:07, 11 October 2007 (UTC)


 * This would be a nice way to handle it. Opus33 20:42, 12 October 2007 (UTC)
 * Ask in the village pump. That's a modification to MediaWiki:Common.js. Tito xd (?!? - cool stuff) 23:26, 27 October 2007 (UTC)

This also encourages two other bad habits:
 * citation of book only, without page number, which is contrary to WP:CITE (because if you cite the book only, you get to repeat the footnote.
 * Dividing a footnote which would say, in almost any printed text: "Source A p. 97, Source B p.238, Source C p.96 n.1." into three footnotes, which makes the text and the list of notes less legible. Septentrionalis PMAnderson 03:00, 28 October 2007 (UTC)

I don't know what this is trying to say, but I don't think I agree:
 * Please note that named references should only be used when there are several cases of repetition of exactly the same reference, down to the page number. Multiple references do have a cost: they make it more difficult for a reader to go back to his place in the article, when several places in text link to the same note. It is perfectly acceptable to simply repeat a footnote instead of using a named reference.

Repeating footnotes will chunk up the article size to an unbearable level, and all you have to do is hit the 'back' button to get back to your place in the article (even for editors who are "her", not a "him"). I think this text is wrong, but it certainly doesn't need to say "him". It also shouldn't say "down to the page number" because websources don't usually have page numbers, and they are a frequent use for named refs. Besides that inaccuracy in difficulty in returning, can someone clarify Opus33's original concern, because I'm not seeing it. The only time named refs are a problem is when editors use them incorrectly to avoid specifying page numbers, but I'm not sure we have a guideline anywhere that demands page numbers on book sources; can anyone point to one? Sandy Georgia (Talk) 03:37, 28 October 2007 (UTC)


 * WP:CITE says full citations for books "typically" include page numbers. Anyway, the original issue is that if a reader clicks on a footnote [1], and the footnote has many backreferences, it's not clear from the footnote alone which letter is the right backreference. But as you say, the browser's back button works. Gimmetrow 05:24, 28 October 2007 (UTC)

Reversions
PBS, you have been reverting one section to a non-consensus version for months. That you are blindly reverting and undoing collateral edits does not help your case. Please cease and desist. Gimmetrow 01:54, 27 October 2007 (UTC)
 * I am reverting to a version that is a compromise that was worked out between a number of editors who hold different opinions. It is you who is reverting to a version which places a prescriotion on other editors for which there is not a consensus. As for your allegation "undoing collateral edits" you are referring to this edit and I do not consider altering the name of a section under dispute back to "Where to place reference tags" from ""Where to place ref tags" to be undoing collateral edits as changing that section heading had broken at least one redirect to this section. --Philip Baird Shearer 11:47, 27 October 2007 (UTC)


 * 1) You changed the name of this section when you began this crusade. I pointed out the problem with the title then, and you agreed to the change then.
 * 2) Please demonstrate that you have consensus for your version, or stop reverting to it. Gimmetrow 13:15, 27 October 2007 (UTC)


 * PBS, please remember that continuing to revert to a non-consensual version, even if not exceeding 3 edits in 24 hours, can be considered edit warring. Sandy Georgia  (Talk) 02:00, 27 October 2007 (UTC)

Sandy please read these sections for previous discussions on this subject:
 * Wikipedia_talk:Footnotes/archive3 (March 2006)
 * Wikipedia_talk:Footnotes/archive3 (April 2006)
 * Wikipedia_talk:Footnotes/archive4 (May 2006)
 * Wikipedia_talk:Footnotes/archive4 (May 2006)
 * Wikipedia_talk:Footnotes/archive5 (May 2006)
 * Wikipedia talk:Footnotes/archive5 (July 2006)
 * Wikipedia talk:Footnotes/archive6 (March 2007)
 * Wikipedia talk:Footnotes (June 2007)
 * Wikipedia talk:Footnotes (July 2007)
 * Wikipedia talk:Footnotes (July 2007)
 * Wikipedia talk:Footnotes (September 2007)


 * Wikipedia talk:Cite sources/archive14 (October 2006)
 * Wikipedia talk:Cite sources/archive16 (May 2007)
 * Wikipedia talk:Cite sources/archive17 (June 2007)
 * Wikipedia talk:Cite sources/archive18 (July 2007)
 * Wikipedia talk:Cite sources/archive18(July 2007)

and then decide which is the "non-consensual version" and which is compromise version. --Philip Baird Shearer 11:47, 27 October 2007 (UTC)


 * Please demonstrate that your version has consensus, or stop reverting to it. That is all. Gimmetrow 13:15, 27 October 2007 (UTC)

It is not "my version" this particular version was a compromise by User:Viriditas of a compromise worked on by a lot of other people. I obviously can not show you, Gimmetrow, that there is a consensus for this version as you do not agree with it. However there is no consensus for the prescriptive version that you insist on inserting into the article. Surly it is better to have a compromise version than a prescriptive version? Why is it that you wish to force one particular style on editors when there is not consensus to do so? --Philip Baird Shearer 14:59, 27 October 2007 (UTC)


 * Don't call me surly. Either demonstrate consensus for your edit, or stop making it. This is no longer a question of "compromise" or whatever. That's been tried and still you insist on reverting to your particular version. Gimmetrow 15:23, 27 October 2007 (UTC)
 * As WP:CONSENSUS notes, both sides in a dispute will claim consensus, in good faith. We've held a poll about this, and the results are plain: there is not now any consensus on either position, that footnotes should go after punctuation or that they should not. There is no consensus (in fact, there is no majority) that there should be a house style at all. There a few things which are not in dispute, and the texts now competing agree on them; on the rest there is none.   The best thing to do in such situations is to state the fact that there is a disagreement. Septentrionalis PMAnderson 02:07, 28 October 2007 (UTC)


 * There are many things which might be agreed to, but PBS has chosen to disregard them and repeatedly reinsert a text which is clearly rejected. That is the point being discussed here. Gimmetrow 04:44, 28 October 2007 (UTC)
 * I agree with Septentrionalis, there is no consensus; neither for the current version, nor for the version inserted by Philip Baird Shearer. Any party changing/reverting changes using the wording 'consensus' is plainly wrong. There maybe other reasons for change, but consensus is not one of them Arnoutf 10:20, 28 October 2007 (UTC)
 * Then why does PBS continue to revert in a text which is clearly rejected? I see he has done it again, again without any discussion on this page, and again without any attempt to find a alternate text which might be acceptable. Others have tried to find alternative texts. I believe there may be no choice but escalation to dispute resolution. Gimmetrow 18:00, 28 October 2007 (UTC)
 * It was the last version to which others agreed without reverting out all the changes. This particular version was a compromise by User:Viriditas of a compromise worked on by a lot of other people. I thought that I had already made that clear on this page. I did not realise that I had to make such a comment every time I edited the guideline.--Philip Baird Shearer 21:08, 28 October 2007 (UTC)
 * That is why it is no consensus but a conflict. I think dispute resolution is not an escalation to what is happening now. Many of the stages described there are aimed at DE-escalating. Arnoutf 18:06, 28 October 2007 (UTC)

I've removed one of the most objectionable parts. Let's see how this goes. Note that according to the poll, the part I removed had 10 disagrees to 6 agrees. I see no reason to add something with less than 40% approval. Since when do we change guideline pages with such weak support? Note that when the "no space" rule was incorporated into the ref-after-punctuation rule, *all objections* were dealt with at the time, so it had 100% support. Gimmetrow 18:30, 28 October 2007 (UTC)


 * Gimmetrow Your change removes ", and editing solely to change from one style to another throughout an article is deprecated unless there is a consensus to make the change." Why is that objectionable? Is it not an invitation to start edit wars between styles? --Philip Baird Shearer 21:13, 28 October 2007 (UTC)


 * It is objectionable because it's disruptive to gnome editing. Do we usually add text with less than 40% support to policies and guidelines? Gimmetrow 23:40, 28 October 2007 (UTC)


 * what is gnome editing? As there is no consensus on a preferred style at a global level then it is better that it is emphasised that there must be consensus at the local level (as is done in the MOS for National forms of English (WP:MOS). The reason it is necessary is as part of a compromise otherwise the "Wikipedia's house style style ..."  has to go, and the implied inbalance in favour of CMS "many editors" and "some editors" needs to be changed to "some editors" and "other editors". --Philip Baird Shearer 07:43, 29 October 2007 (UTC)
 * Why is it so difficult for us to compromise? Let's just state the facts and be done with it. What do we know so far? &mdash;Viriditas | Talk 08:34, 29 October 2007 (UTC)
 * Gnome editing on this detail should be discouraged; just as the occasional editor who has attempted to "clean up" the color/colour inconsistency should be discouraged. In one case, it's a difference of national variety; in the other, it is a difference of preference which may be influenced by nationality. But in both cases, we should learn to leave things alone. Septentrionalis PMAnderson 18:43, 29 October 2007 (UTC)
 * Wow. I was not aware you rejected ENGVAR too. In regard to your statement that articles looking like crap should be left alone, I'm not sure how to argue with that, other than to say I really doubt many editors agree with it. Perhaps you should learn to leave editors alone fixing articles? Gimmetrow 14:44, 30 October 2007 (UTC)
 * Please have the civility to read what I actually said; I support ENGVAR, and think its principles extend here. (It is possible that my example has confused you: I object to the editor who takes an article which always uses color or colour and "corrects" to the "right" spelling; do you really mean to say that an article which consistently places its note marks before punctuation looks like crap?) Septentrionalis PMAnderson 17:18, 30 October 2007 (UTC)
 * Please also read what I actually said, too. Your reply in that context says if an article uses both "color" and "colour", that editors should be forbidden from making the article consistent. Gimmetrow 17:37, 30 October 2007 (UTC)

PBS has again inserted a text with less than 40% approval, per the poll. This is unacceptable. Is anyone else willing to move to a user conduct RFC? Gimmetrow 13:32, 29 October 2007 (UTC)
 * Text in question: ...and editing solely to change from one style to another throughout an article is deprecated unless there is a consensus to make the change. That seems pretty silly.  Editors change citation style all the time, without establishing consensus.  Perhaps it could be worded differently. &mdash;Viriditas | Talk 13:51, 29 October 2007 (UTC)
 * Less than 40% or more than 40% it makes no difference. Polls are evil. Try to understand the arguments, and then you can counter them. Refusing to reason because of "majority rule" is not acceptable. --Mig77(t) 13:55, 29 October 2007 (UTC)
 * It's come to this precisely because someone was continuously reverting without discussion. You can't discuss if the other party refuses. Gimmetrow 14:07, 29 October 2007 (UTC)
 * A conversation involves two. Please answer the questions I have asked you instead of just making statements like "PBS has again inserted a text with less than 40% approval, per the poll. For example:
 * what is gnome editing? As there is no consensus on a preferred style at a global level then it is better that it is emphasised that there must be consensus at the local level (as is done in the MOS for National forms of English (WP:MOS). The reason it is necessary is as part of a compromise otherwise the "Wikipedia's house style style ..."  has to go, and the implied inbalance in favour of CMS "many editors" and "some editors" needs to be changed to "some editors" and "other editors". --Philip Baird Shearer 07:43, 29 October 2007 (UTC)
 * ---Philip Baird Shearer 15:13, 29 October 2007 (UTC)


 * See WP:GNOME. This line is disruptive; it would require editors to ask on the talk page "Please, may I change 5 of the 16 references which appear to be inconsistent to make them consistent", then wait 2 weeks to see there is no objection, otherwise enter into a lengthy discussion on each and every page over what should or shouldn't be the style for that page. That's an absurd process. In fact, that's one reason why style guides exists, so we don't have to have the same discussion ad nausem. Wikipedia has decided not to use the double hyphen--and also not to use "American style quote-punctuation." Do you wish to dispute those decisions as well?
 * I do not see this as comparable to national styles of English. Nobody has provided a single style guide suggesting that a national style is involved. And in any event, why would the mention of "house style" have to go, and why would the text not recognize the preferential practice, even if it were to allow a minority practice? Finally, in the interest of "answering questions":
 * Smith lived in Canada in the 1950s. He was six foot six tall and he weighed 225 pounds, but that giant of a man was smitten by love[2].
 * How can you tell, from the text alone, that [2] refers to the last phrase, the last sentence, or both sentences? Gimmetrow 16:28, 29 October 2007 (UTC)
 * that argument is spurious as the current guideline would say the footnote is always after the period, so no way to distinguish either. Arnoutf 18:05, 29 October 2007 (UTC)
 * PBS claimed "Of course you can tell". I asked him to explain. I would like an adequate response. Gimmetrow 00:47, 30 October 2007 (UTC)
 * The reference to "house style" was inserted without discusssion by a single editor, and has never had consensus. Some editors have agreed with it (and for my part I support saying that some editors do), but that's not the same thing. Septentrionalis PMAnderson 18:46, 29 October 2007 (UTC)
 * Gimmetrow, I don't think that is how an edit would work. Let us suppose that the wording is "House style, CMS. also Nature style," + "editing solely to change from one style to another throughout an article is deprecated unless there is a consensus to make the change". An editor who wished to change from Nature to House style, would make the change and if no one objected then the edit would stand. But if an editor reverts the changes and says I prefer Nature, this clause stops edit wars. Without that clause and a house style, some editors will try to browbeat editors who prefer the Nature style to accept the house style even thought all those involved in this debate know there never has been a consensus over this issue. If we remove the house style and use   "some editors" and "other editors" then there is no need to have the consensus sentence as the default will be whatever the people who edit the page want to use. --Philip Baird Shearer 18:54, 29 October 2007 (UTC)
 * To answer you question I would take [2] to mean the last sentence, but I am not expert on the Nature style, far better to address that to a person who reads Journals such as Nature regularly. --Philip Baird Shearer 18:54, 29 October 2007 (UTC)
 * Better yet, ask someone who publishes in it regularly ;-) But no kidding, this is the nature style (from its own webpage) Arnoutf 19:01, 29 October 2007 (UTC)


 * Are you disputing that a majority of editors use the ref after punctuation style? It has, after all, been discussed and supported for over 2 years, and continues to be the house style, and according to the evil poll, a substantial majority say it is the house style. All the parts of the original text were discussed and added without, as far as I recall, any standing objection. Putting the house style on an equal footing with a minority style, one which I don't think even PBS likes, seems rather difficult to justify. I also see no need for the additional line, as this seems hardly comparable to national styles of English. Again, Wikipedia has a rule against "American" or "typesetter's quotation," and nobody seems to be pressing "national style" over that. I would also appreciate it if the project page were not continuously edited live, which was the point of this thread. Gimmetrow 00:47, 30 October 2007 (UTC)
 * This is wrong on both points. The absence of "typesetter's" or "aesthetic punctuation" has been questioned by an editor, new to the discussion, about once a month since I've been watching WT:MOS; at least half the time, this is an American requesting equal rights for Americans. (Whether this is true is another question; I can testify that many Americans are taught it, but I do not use it myself.) Septentrionalis PMAnderson 04:13, 7 November 2007 (UTC)
 * Comment above added over a week after this discussion. What happens when this occasional editor "questions" the absence of typesetter's quotation? Does this point get "pressed" (the word I used)? Does it overturn the existing MoS rule/consensus? Should it? Gimmetrow 00:15, 8 November 2007 (UTC)
 * How many active wikipedia editors are there? I do not think that the very small participation in a survey, even compared to the number of editors who have participated in this debate since it was initially raised in March 2006, can in any way be considered an indication that "a majority of editors use the ref after punctuation style". I suspect the vast majority of editors do not use reference tags at all. Even if it were true, that a majority do, that does not mean that there is a consensus to impose a prescriptive style. If we are going to have a compromise along the lines of "Frequently, a reference tag will coincide with punctuation and many editors put the reference tags after punctuation (except dashes), as is recommended by the Chicago Manual of Style (CMoS). Some editors prefer the style of journals such as Nature which place references before punctuation." Then I think it is necessary to have "Each article should be internally consistent, but editing solely to change from one style to another throughout an article is deprecated unless there is a consensus to make the change." to stop edit-waring, just as we have for other areas of the guidelines were there is no general consensus on an issue. --Philip Baird Shearer 14:23, 30 October 2007 (UTC)
 * I see no reason for anything with the word "deprecated". Other than project pages, what edit wars have there been over this issue? Gimmetrow 14:44, 30 October 2007 (UTC)
 * There have indeed been edit wars over this issue; I don't at the moment remember which pages, since I tend to avoid such sophomorism. I don't see why it should not be deprecated. Septentrionalis PMAnderson 17:18, 30 October 2007 (UTC)

How many editors read and approved the text as it developed over two years, PBS? What *exactly* are the problems with the prescription that absolutely must be addressed by the removing the prescription and allowing a second style? *If* the project page were to allow for two styles, and I still see no reason it needs to, what exactly will that accomplish? Is that what you really want, or is this a step to something more? Gimmetrow 14:44, 30 October 2007 (UTC)
 * The 'one-style-only' is a contested restriction on editing freedom. That much is clear. I would say, in case of a conflict where there is no consensus or at least overwhelming majority (40-60 as Gimmetrow says is not a large majority) I would be against restricting guidelines. This is why I would favour allowing both styles purely on procedural grounds. Arnoutf 17:44, 30 October 2007 (UTC)

In response to an earlier question, I don't think a RfC/U is the next best step here (besides that RfC/U is a highly abused process because it doesn't attract a wide base of uninolved editors, this is more of a content dispute than user conduct). I do think this needs to be exposed to a wider audience and settled. The consensus stood for many years, a minority challenged it, and it needs to be resolved with broader input. Has this gone to the Village Pump, for example? As I seem to recall many argued earlier, expecting Wiki to follow the outside style of one publication (Nature) is no longer realistic for a venture that has reached a level that warrants having our own style. As Gimmetrow has pointed out, "gnomes" should be able to contribute to maintenance of Wiki articles, so having an inhouse style facilitates article improvement. This is not the same argument as ENGVAR. I support a return to the long-standing consensus before the minority (Nature) challenge: refs after punctuation. Sandy Georgia (Talk) 17:52, 30 October 2007 (UTC)


 * SandyGeorgia if you read the contents of the links I put at the top of this section how can you write "the long-standing consensus" because AFAICT, (perhaps apart from a short while when a compromise was agreed this summer), there never was a consensus? "expecting Wiki to follow the outside style of one publication (Nature) is no longer realistic" yes I agree, but no one who is suggesting a compromise is suggesting one outside style from one publication, it is those who want a prescription who want to use the one outside style from one publication (CMS). "so having an inhouse style facilitates article improvement" how does CMS style over any other arbitrary third party style "facilitates article improvement"?  --Philip Baird Shearer 01:39, 31 October 2007 (UTC)


 * There was no consensus in the summer discussion. There was a discussion, a semi-tolerable text was developing, and the discussion was derailed. There was never a consensus. Quit repeating that. What you are proposing is that WP have no in-house style at all. I don't really care what the style is, but I think there needs to be one so people can move on and not write megabytes discussing this point. Most points of the MoS were decided based on the majority practice in writing in a variety of fields, which also maintains a "principle of least surprise", in that generally the styles suggested by the MoS will not be unusual or difficult to follow for most readers. There are quite a few details of the MoS that I would rarely if ever use in my own writing outside WP, but they are the chosen style here. Gimmetrow 02:18, 31 October 2007 (UTC)


 * Gnomes would be useful if they made the style of any given article consistent. We have no need for a house style, especially if the purpose is to show we are all grown up now; the test for that is the quality of our articles, not their punctuation [style 18:26, 30 October 2007 (UTC)].


 * No one has argued that editors must follow Nature's style; merely that editors who prefer it should be free to do so, either in new articles or by consensus in existing ones. I regret the presence of this red herring. Septentrionalis PMAnderson 18:01, 30 October 2007 (UTC)


 * I regret that you consider that a venture of the size and reach of Wiki doesn't warrant an inhouse style, and that you express that article punctuation isn't a component of our quality: I (and others) believe it is, and would not classify it as a "red herring".  Sandy Georgia  (Talk) 18:09, 30 October 2007 (UTC)
 * I regret that Sandy has misunderstood what I called, and still call, a red herring: her straw man argument that someone, anyone, wants to require that we use only Nature 's style, which is fact widespread in the scientific literature. Her account of the course of this dispute is equally inaccurate: the "house style" was declared by a single editor, two years ago, and has been protested ever since.


 * I have emended, above, another word that may have given rise to genuine misunderstanding. Septentrionalis PMAnderson 18:26, 30 October 2007 (UTC)

Straw men
No one has ever said that Wikipedia is too small to "warrant" a house style; there are merely points on which having a mandatory style would not serve us. One is ENGVAR; another, even closer to this discussion, is the question of whether an article uses Harvard referencing or footnotes in the first place. Editors are free to use Harvard referencing on new articles; a consensus of editors can introduce it; but any given article should use one or the other. So here; an editor that prefers after punctuation should be free to begin an article using it, or a consensus of editors can introduce it; the same thing with before punctuation. Septentrionalis PMAnderson 18:34, 30 October 2007 (UTC)

Likewise, no-one wants to require that we use only Nature 's style, which is fact widespread in the scientific literature; we merely want to permit it. Septentrionalis PMAnderson 18:47, 30 October 2007 (UTC)


 * Yes, please be aware that this is part of a continuing war by a few people against the notion of centralised guidance, even rules, to bind WP into an authoritative, cohesive project. Here, someone is beating a worn-out drum for the so-called illogical punctuation that Chicago persists with. This mantra should be resisted because it goes against an overarching principle on WP of leaving sources be. (I don't want to argue further about that here, though—we're all sick of it.) Similarly, where there are good reasons to enforce a single style, this should be done. I believe that the before-punctuation style looks ungainly and disjointed—harder to read—and when there's a string of numbers, it's downright ridiculous. As for enforcing logical punctuation, post-punctuation references should be chosen, I believe. WP is no longer an adolescent, and should adjust to adulthood. Please let's not kow-tow to other authorities for the sake of it, or shirk our responsibility to maintain good guidelines. Tony   (talk)  02:31, 31 October 2007 (UTC)
 * It is Wikipedia's misfortune that a few provincial and utopian editors would like to use it to create a new, "improved", "modernized" English. Fortunately, these ideas do not have, and never have had, consensus. Septentrionalis PMAnderson 14:57, 31 October 2007 (UTC)
 * What in the world are you talking about? Here, the WP house style on footnote placement is a style which is generally thought to be the more common and least unusual usage in a wide variety of fields. I have seen no suggestion that it's part of some utopian vision of modernized English. I would further question whether the Nature style can really be called "widespread" even in scientific literature; most scientific literature uses Harvard citations or some variation, including various journals in the Nature Publishing Group. Among those journals where footnotes are used in article text, ref marks are probably most often placed after punctuation, also including some journals in the Nature Publishing Group. Gimmetrow 15:52, 31 October 2007 (UTC)
 * For an example of utopianism, see this effort to legislate a new "logical" English; this is one example of all too many. (I acknowledge freely that Gimmetrow has nothing to do with this; the merits of punctuation after or before footnotes is another question, into which the ideologues are intruding.) Septentrionalis PMAnderson 16:34, 31 October 2007 (UTC)
 * I see no evidence above on which preference on placement is more common, and therefore I have attempted to word so as to say nothing on the matter. Septentrionalis PMAnderson 16:35, 31 October 2007 (UTC)

Per PMA's html comment to the aritcle that "evidence should be placed on Talk", the "evidence" has been on talk, and you have been so directed there already. I refuse to put dubious and disputed and other absurd tags on guideline pages. Gimmetrow 22:56, 5 November 2007 (UTC)
 * False, both on what been said here, and what has been shown. The poll divided evenly on this matter; and no evidence has been presented otherwise. Guidelines should not, in general, make claims of fact, and I don't see why this one should; and this claim is both dubious and disputed. As long as it is insisted on, it deserves to be tagged. Septentrionalis PMAnderson 23:42, 5 November 2007 (UTC)
 * The poll did not divide evenly on the matter. QED. Gimmetrow 00:07, 6 November 2007 (UTC)

The poll is such a small sample that any claims about the number of editors based on that poll fails statistical sampling methods. --Philip Baird Shearer 12:35, 6 November 2007 (UTC)
 * I partially disagree with PBS about sampling; I (partially) agree with his conclusion, but not his explanation. Statistical tests take account of sample sizes, for example a division of 4000-6000 is likely an indicator of a true difference and will give a statistically significant deviation from 50-50 split. This will probably also be the case for 400-600. It is no longer certain 40-60 is statistically different, and 4-6 is most definitely not. Hence QED conclusion that the poll did not divide evenly is not proven (ie there is no significant deviation from a 50-50 opinion). Whether this is because of the small sample size; or because there is truly no difference is not decided, however the poll does not give an indication of anything but even division (as PManderson correctly remarked). The poll sample is however more fundamentally flawed, in that the poll was not randomly chosen among all wiki editors, but consisted of self-selected participants. This is likely to introduce bias (people with the strongest feelings are more likely to participate), and hence this is the main reason why these polls cannot be generalised. Arnoutf 12:47, 6 November 2007 (UTC)
 * "the poll does not give an indication of anything but even division (as PManderson correctly remarked)": No, it does not "indicate" even division in any positive sense, though from a statistical standpoint it may not be a large enough sample to rule out a purely 50-50 distribution with any certainty. However, going only by statistically valid conclusions, one also cannot conclude from the poll anything about consensus, including that it shows sufficient consensus to change from the established text of this page. The sword cuts two ways. Gimmetrow 02:51, 7 November 2007 (UTC)
 * I have been consistently attempting to avoid making any quantitative claims on this evidence; this is why I support "some"/"some" (which says to a mathematician: "not zero" on either side). But "one party"/"the other party" may be an improvement. Comments? Septentrionalis PMAnderson 14:32, 6 November 2007 (UTC)

You want the page to allow two styles. Fine, I completely disagree with that, but PBS has carried this on long enough to get squeaky wheel consideration, and the page can tolerate exactly two, and no more than two, styles if that will remove the squeak (get the page stable). Still, these styles are not equal, neither among journals, nor among WP editors. Clearly one of the two styles has precedence in time on this page, and that needs to be recognized by this page. I frankly find no reason for you to object to this, since this same statement is made on another page which IS stable. (I suppose that, because I've mentioned it, that page will now be subject to this silliness.) Gimmetrow 02:51, 7 November 2007 (UTC)
 * This demonstrates that Gimmetrow has a completely different fundamental philosophy than ours; and one, I think, demonstrably contrary to policy.


 * This is a flat denial of WP:Consensus can change. We do not need to recognize what we used to do; Wikipedia is not bound by precedent. That's one reason we have edit histories: so we can find out what we used to do while now doing something completely different.


 * In this case, furthermore, the only recognition deserved by the unilateral and consistently disputed declaration by a certain disruptive and revert-warring admin (and I mean none of those presently conversing) is as evidence in the eventual ArbCom case.


 * However, in a more peaceable mood, let's see what text Gimmetrow proposes; if I can tolerate it, I will not revert. Septentrionalis PMAnderson 03:37, 7 November 2007 (UTC)


 * I could live with the WP:CITE text, although I disagree with it. And no, this is not a denial that consensus *can* change, it's a denial that consensus *has* changed. I think most editors are fine with the house style as it was, regardless of how it got there. I would expect any tolerable text will say there are different styles, and that WP recommends but does not require the ref-after-punctuation style. And I would continue to object to any phrase with "deprecate" that appears to be disruptive to article cleanup. Gimmetrow 13:55, 7 November 2007 (UTC)
 * That is clearly what Gimmetrow believes. I join with the several comments above which deny that a 6-4 straw poll cen be a sound justification for that belief. Septentrionalis PMAnderson 19:20, 7 November 2007 (UTC)
 * 6-4? Apparently we're not looking at the same poll. But are you really claiming that 4 votes can be a sound justification to change this page? Gimmetrow 00:15, 8 November 2007 (UTC)
 * Why not, if one vote was enough to install the "house style" in the first place? :}


 * But the serious answer is: Yes, certainly; the poll is sufficient evidence that we are divided into two roughly equal camps. The claim that the CMS rule is consensus is therefore false, and should be removed; we should not claim consensus where it does not exist. Septentrionalis PMAnderson 03:10, 8 November 2007 (UTC)

The reason why WP:CITE has not been kept in line with this page is that we may as well discuss it in one place and the banner at the top of the relevant WP:CITE section that says "Further information: Footnotes" I read to mean the details of the guideline are on this page. Once this page is stable we can alter the cite page to reflect the consensus on this page. --Philip Baird Shearer 08:15, 7 November 2007 (UTC)

Gimmetrow you write "And I would continue to object to any phrase with "deprecate" that appears to be disruptive to article cleanup." "deprecate" is not disruptive to article cleanup, "deprecate" is there to discourage disruptive "cleanups". If there is a bias towards anyone particular style in this section, then without this final sentence there will be people (acting in good faith) who will try to change the style used because they will see themselves as helping the project by changing to the house-style using arguments as we have seen on this page that it is better for Wikipedia if there is one universal style. --Philip Baird Shearer 14:49, 7 November 2007 (UTC)
 * No, you do not need that final sentence. There are many ways to say much the same thing without being disruptive to article cleanup. For instance, you could say: WP recommends A, but B is also acceptable. This formulation doesn't forbid an editor from making an article with mixed A and B consistent. If you do not find that acceptable, why not? Gimmetrow 00:15, 8 November 2007 (UTC)
 * Because some will take that as a licence to change from the acceptable to the recommend. The wording "but editing solely to change from one style to another throughout an article" does not prohibit an editor making an article consistent (note the word solely) --Philip Baird Shearer 12:41, 8 November 2007 (UTC)
 * If there is an article with 8 instances of A and 2 of B, cleanup to have 10 instances of A involves changing all instances of B to A. That sounds to me exactly like "editing solely to change from one style to another throughout the article", which is why I have a problem with this formulation. On the other hand, if A is recommended but B is permitted, and someone tries to change an article with 10 B and no A, citing a recommendation, just point out to the editor that B is also allowed so it's not necessary to change. Is there some other way to say this? Gimmetrow 14:12, 8 November 2007 (UTC)
 * The intention of the wording was to have the effect of MOS: Retaining the existing variety how about:
 * 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 for changing it.
 * --Philip Baird Shearer 04:30, 10 November 2007 (UTC)


 * I don't see this as an issue of regional variety of English. I've looked for a style guide suggesting something opposed to ref marks after punctuation, and haven't found one. Yes, some journals have their own forms, but those are requirements *to publish in that journal*; likewise for internal style guides. General style guides appear pretty uniform in saying ref marks go after punctuation. We already have CMoS. The MLA Style Handbook says: "They follow punctuation marks except dashes and occasionally parenthesis. (When the note is to only the material that appears within parentheses, the note number is placed before the closing parenthesis.)" (1985, p. 189). I could cite others, but they are generally based on either CMoS or MLA. Many less-comprehensive style guides don't explicitly discuss the details with footnotes, but they show ref marks after punctuation in every example. Gimmetrow 17:19, 14 November 2007 (UTC)
 * You are now putting forward a different argument, can we please keep to the issue about the wording and not try to solve everything at once. I suggest that we put the issue of different style guides to one side in this thread for the moment and concentrate on the issue you raised about the interpretation of the sentence "Each article should be internally consistent, but editing solely to change from one style to another throughout an article is deprecated unless there is a consensus to make the change.". Does the alternative wording "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 for changing it." meet your concerns about homogenising all the references to a particular style in a specific article? --Philip Baird Shearer 19:12, 14 November 2007 (UTC)
 * Yes, that language solves the one problem about homogenising, but it gets back to the main issue, which is that I don't see these two styles as equal. One is common and recommended by "many" style guides (when footnotes are used), and the other is used by "some" journals. Using the language of a first predominant style seems inappropriate. I would prefer something like: "Place a ref tag at the end of the term, phrase, sentence of paragraph to which the note refers. When it coincides with punctuation, Wikipedia's recommended style is to place the ref tag directly after the punctuation mark without an intervening space. This is the format recommended by the Chicago Manual of Style and the MLA Handbook. Nevertheless, guidelines should be treated with common sense. Some editors prefer the style used by Nature, which places ref marks before punctuation; if the editors of an article have a consensus to use that style, the style should not be changed." Gimmetrow 22:57, 14 November 2007 (UTC)
 * Leaving aside the argument over style guides and concentrating on the consensus issue. I do not think that you new wording is acceptable because the person wishing to change from one style to the other will argue that as they want to change it there is no consensus for the current style so they are free to change it. ---Philip Baird Shearer 12:23, 15 November 2007 (UTC)
 * But one person (or a small number) wishing to change does not overcome an existing consensus. I want this to state one recommended style, the style which has been discussed and built upon on this page for more than a year by dozens of editors. However, this *is* a guideline, and guidelines admit exceptions, although exceptions should be based on something more than an individual preference. Thus if editors at an article agree to use Nature style in a science article, that's fine. Gimmetrow 19:35, 15 November 2007 (UTC)
 * There is disagreement over whether there has ever been consensus for wording/style/prescription that SlimVirgin inserted into the guidelines without building a consensus to do so. The archive sections listed at the start of this section clearly show that there never has been a consensus for this and we are still discussing it. I was merely trying to remove one of the disagreements that you and I have had over changing styles without consensus sentence. As to your wish to have a house style, this I think is a question about whether the word "some" or "many" is inserted in this sentence "Frequently, a reference tag will coincide with punctuation and [many/some] editors...". I personally am willing to acquiesce to the word "many" providing the sentence containing "..predominantly one style of ref tag placement.." is included to protect those articles which use before the punctuation. This is a compromise around which I hope we can build a consensus even if we do not get all we want. --Philip Baird Shearer 22:34, 15 November 2007 (UTC)

discussion on the content of the text

 * This is a continuation of the previous section and should be read in that context.

It doesn't matter how the text got there initially. Dozens of editors have worked on the text discussing the house style; building on and developing it is a demonstration of consensus. Yes, consensus can change, but the basic problem I have is: the former text was thrown out by a handful of editors, far less than the number who have participated even in small parts of developing the old text. If this new handful will not respect the past consensus, then I hope they have no expectation the new text will be respected by anyone. My solution is to say WP recommends the style it always has, with an indication that guidelines admit exceptions. Gimmetrow 23:34, 15 November 2007 (UTC)

I have had a look at the history of the article before 17 May 2006, when SlimVirgin introduced her prohibition the previous wording was in  section called "Style recommendations"
 * Citations should always follow punctuation,[5] like this.[5]

The history of the prohibition is: Section "How to use:"
 * 05:38, 17 May 2006 SlimVirgin
 * Introduced the prohibition with no discussion on the talk page:
 * Note that the ref tag comes after a period (full stop) or comma, and a space should be left between the period and the tag. Do not place the period after the tag. A sentence should be written like this. It should not be written like this .

Only 5 other editors worked on the section over the next month (one of whom was against the prohibition, one made edits supporting it, and the other three did not edit for or against it but edited another sentence):
 * 07:17, 17 May 2006 Ligulem
 * 15:28, 20 May 2006 Circeus
 * 17:54, 22 May 2006 Xoloz
 * 18:21, 22 May 2006 Francis Schonken
 * 04:28, 24 May 2006 Mbeychok
 * 09:16, 14 June 2006 Ligulem
 * Introduction of first compromise wording:
 * The basic concept of the tag is that it inserts the text enclosed by the ref tags as a footnote in a designated section, which you indicate with the placeholder tag . Note that many editors put the ref tag after a period (full stop) or comma (if it refers to the whole sentence), which is also recommended by the Chicago Manual of Style (CMoS), but there is no consensus to have it like that on Wikipedia. ...

Since June 2006 there have been editors editing the page on both sides of the argument. So the prohibition was only in place for one month before a compromise was first edited into the article page after discussions on the talk page -- see the first complaint on the prohibition on the talk page: 3 days after the introduction of the prohibition 08:28, 20 May 2006 Robdurbar

I think that the evidence in the edit history of the article and the talk page shows that there has never been a consensus for a house style such as that introduced by SlimVirgin on the 17 May 2006. --Philip Baird Shearer 11:37, 16 November 2007 (UTC)


 * As I recall, the initial dispute was over the "space" aspect of the rule. This was immediately brought up in discussion May 17, 2006 ( copy). When this was discussed in  August 2006, agreement was for no-space, and nobody was disputing the location of refs in relation to punctuation. When this was revisited again in  March 2007, quite a few editors commented; discussion was heavily for no-space, and again nobody said anything against having refs after punctuation. The text recommending refs after punctuation stayed in place for over a year until it was removed based on an aborted discussion at a different page. The former text here was developed and built upon. I think that the evidence shows that there is a consensus for a house style. Gimmetrow 13:54, 16 November 2007 (UTC)

The space debate started on the 17 May the first complaint about the position of the footnote was three days later (08:28, 20 May 2006 Robdurbar ) see "Wikipedia talk:Footnotes/archive4" which was only three days after SV's prescription was added to the page! If you look at the start of the parent section above you will see that this issue about placement before or after punctuation came up several more times before March 2007! And as I have pointed out the first edit to the page with a compromise was on 09:16, 14 June 2006 (Ligulem) less than a month after SV's prescription. The last time I made an edit over this issue on this page in 2006 was on 18 July 2006 because there seemed little point edit waring about it (as everyone had made their POV clear) at that time, but the issue came up several more times on the talk page. So are you suggesting that a change to policy is accepted even if discussion continues on the talk page unless there is an active edit war on the policy page? Philip Baird Shearer (talk) 17:41, 16 November 2007 (UTC)
 * Um, not exactly. I'm suggesting that if a text is stable for a long time, including building on an allegedly disputed point of that text without opposition, the dispute isn't enough to overturn consensus. Consensus does not mean 100% positive agreement. If the same issue is raised five times, and only on the fifth time gets a few editors in support, what has changed? Consensus? or merely the people involved the fifth time? It doesn't help when the discussion starts on another page, giving the appearance of "forum-shopping". Gimmetrow 18:20, 16 November 2007 (UTC)

A survey of four weeks of Main Page articles
I've been puzzling over the use of "Notes", "Footnotes", "References", "Notes and References", etc., as titles for sections at the bottom of articles, so I thought I'd look at Main Page articles for the past four weeks.

Survey results
None of the 28 articles used Harvard citation templates; all used the cite.php style footnotes. However, a number used Harvard-citation style footnotes in whole or part; I'm going to call this "academic" style footnotes here. The split was 5 articles with academic-only, 11 mixed, and 12 full-citation (what I'll call "regular") footnotes only. For the middle category, "mixed" style, there were actually two variants: either two sections (one for footnotes, once for sources, as is the case with Harvard citations) or a single section (the first time a footnote cites a source, full information is given; for the second and subsequent footnotes citing that source, only the author and page number is given).


 * Five articles used a "pure" academic approach: footnotes in the first section, citing author and page number, and full citations in the second section. (Example: Donkey Kong (video game). The second section was always called "References", but the name of the first varied:
 * "Citations" - once
 * "Footnotes" - once
 * "Notes" - three times


 * Eight articles used a "mixed" approach with two separate sections. In the first section were two different types of footnotes (some full citations and some just author and page number); the second section had academic-style full citations. (Example: Cardinal-nephew.) The name of the second section was almost always "References" (with Four Times of the Day, it was "Sources", with the footnotes in the section titled "References", which appeared first), but the name of the first section, with the footnotes in it, varied:
 * "Footnotes" - twice
 * "Notes" - four times
 * "Notes and sources" - once


 * Three articles used a "mixed" approach with a single section (the first time a footnote cites a source, full information is given; for the second and subsequent footnotes citing that source, only the author and page number is given). Each of them was different:
 * Bob Meusel had only three "academic" footnotes, most were regular footnotes. The single section was called "Footnotes".
 * 1880 Republican National Convention had its single section called "References"; mixed in were some footnotes that were pure notes (no source cited; interpretative.
 * William Shakespeare also had a single section called "References", but had a separate "Notes" section that used the Note_label template.


 * The twelve articles that used exclusively regular footnotes were the most consistent. Nine did not have notes; in all cases, the footnotes section was titled "References". (Example: Enzyme kinetics.) For the other three, which did have notes (again, I'm defining "notes" as commentaries on the text that don't cite a source):
 * Two had a single section called "Notes and references". (Example: Barnard's Star)
 * One (John Mayer) had the footnotes in a section called "Footnotes" and the notes (which used the cnote template) in a section called "Notes".

Comments on the survey

 * Obviously Wikipedia has a long way to go regarding standardizing footnotes if even Main Page articles have such variation.


 * While I'm not arguing for a single way of doing things, some consistency - two basic models, with standardized variants, for example, might prevent arguments and give readers a sense that this is one encyclopedia, not a bunch of articles worked on by different groups of editors.
 * But that is exactly what Wikipedia is: a bunch of articles worked on by different groups of editors.Septentrionalis PMAnderson

(Note: for discussion of the placement of tags, see the next (main) section; I'm separating subsequent comments into that separate section because they're not on point regarding style. -- John Broughton  (♫♫) 22:52, 16 November 2007 (UTC))
 * The only approach to footnotes that I think is clearly wrong is putting academic citations in a single section with regular footnotes, as with Bob Meusel, William Shakespeare, and 1880 Republican National Convention. When a reader clicks on footnote 79 and sees only " Ackerman (2003), p88.", he/she has no way to figure out that footnote 2 actually has the full citation. -- John Broughton  (♫♫) 23:56, 4 November 2007 (UTC)
 * I agree that this last is a problem; although it is routine in academic publishing to cite books in short form after using them once, whether they are in the bibliography or not. At least we have a search function which will in time find footnote 2 (this is in 1880 Republican National Convention).
 * We could recommend that any book cited twice should be expressly cited in the References outside footnotes. (And Orion (mythology)] uses op.cit. to mark a link to the list of works referenced; this is non-standard, of course, but may be useful elsewhere.) Septentrionalis PMAnderson 23:52, 5 November 2007 (UTC)
 * I'll point out my agreement with PMAnderson that it's odd you feel we should be projecting the image of "one encyclopedia, not a bunch of articles worked on by different groups of editors." Considering that the latter description captures us perfectly, what's the motivation for the deception? Christopher Parham (talk) 01:23, 6 November 2007 (UTC)
 * I don't see anything in PMAnderson's comments referring to "one encyclopedia", so I'll respond only to the last comment. I could have phrased my comment better, perhaps something like this: The purpose of the Manual of Style is to provide guidance to editors so that articles are done consistently; consistency is a help both to readers (finding things) and to editors (creating and improving things). The wide variety of approaches illustrated by the survey I did is therefore counter to the general approach that the project has taken regarding consistency, and is less than helpful to both readers and editors. -- John Broughton  (♫♫) 16:11, 6 November 2007 (UTC)
 * To address the broad level first, I think the project's "general approach" to consistency is disdain. This is an area where the views of the top-100 editors probably differ greatly from the rest of the population which generates the majority of the project's content. (I say this based on my experiences with recent changes and new pages.) In this particular case I don't think that consistency would add much value for readers or editors. PS: the comment of PMA's that I am agreeing with was interjected into your remarks. Christopher Parham (talk) 00:26, 7 November 2007 (UTC)
 * Mr Broughton is correct that I did not use the words "one encyclopedia". I am not sure whether WP is one encyclopedia or not; I'm not entirely sure what that would mean in terms of this discussion. Mr Parham is correct that most contributors don't care about consistency at all. All too many of the remainder have caught on to some accidental trend and are imagining a rule out of it.   Septentrionalis PMAnderson 03:02, 7 November 2007 (UTC)
 * As the author of one of those main page articles, I'd say that standardisation of footnote style would be an excellent aim, and I would advocate in-line cites using reference templates and cite php, with the tags placed after punctuation, as our house style. Tim Vickers 05:14, 8 November 2007 (UTC)