Wikipedia talk:Manual of Style/Archive 124

Citations vs. attributions

Hi, all. I am transporting the following discussion from Wikipedia talk:Manual of Style (words to watch) to here...because not only is this page very active, this discussion signals that a rewrite of a Manual of Style guideline is needed. An editor at Wikipedia talk:Manual of Style (words to watch) stated that we were in the wrong place, when I feel bringing this up there or here is relevant (per the misunderstanding in guideline). The discussion goes as this:

I tagged several sentences in Death of Caylee Anthony for having weasel words, which was quickly reverted by another editor who stated they were not weasel words because they were all properly "attributed". However, I believe the editor is confusing "attributing" with "citing". Here is an example of a sentence that I tagged with {{by whom}}:

"The trial has been compared to the O. J. Simpson murder case, both for its widespread media attention and reported "shock" at the "Not Guilty" verdict.

The tag for this any many other sentences were removed (not in an edit, but rather a careless revert). The citations at the end don't warrant attribution because they need to be in the prose. I have left the editor a message on the talk page, but based on a previous discussion, the editor has had "years of experience" with weasel words and seems to know exactly what he/she is doing. I just want to get some clarification about this to avoid an edit war and to make sure we are all on the same page.–Dream out loud (talk) 17:59, 17 July 2011 (UTC)

As I stated to Dream out loud on my talk page, while I agree that "most," "some" and "others" are weasel words when attributed to what people think or feel or how they behave, I disagree that all these people must be named in the text. As I stated, "The comparison to the O. J. Simpson murder trial is backed up by several reliable sources. I shouldn't have to name each and every one of the individuals who have compared this trial to the O. J. Simpson case, especially since it's too many to name (including the general public)." ...In fact, doing so is never practiced on Wikipedia, and I'll get to that in a moment. Look at the following line quoted by Dream out loud above. In terms of excluding the names of people, how is it any different than saying "The 9/11 attacks had immediate and overwhelming effects upon the American people" (as seen in the 911 attacks#Aftermath section)? Other than the fact that the latter line is specifically mentioned as effecting the American people through text, I'm not seeing a big difference in the style. Are editors here saying that each and every person, regular people included, who have compared this case to the O. J. case should be named? Are editors here saying that "most," "some" and "others" should never be used and that we should list all the names every time? If so, I'm saying that is never done. Not on Wikipedia or anywhere. Otherwise, it would be a namefarm and quite possibly a WP:Linkfarm. This happened with the 2011 Tucson shooting#Political figures section. See where it says "numerous" instead of listing each and every one? For further example, we don't list every celebrity who has spoken out on something. We say "various celebrities" or "several celebrities," etc. Sometimes, it's not even possible to name exactly who all these people are, which is exactly why even reliable sources sometimes state "most Americans." And when even the reliable sources themselves are using "most, "some" or "others," we are simply relaying what the sources are stating. Sometimes, it is a matter of "most," "some" or "others" and we cannot simply leave it as "people."

On a side note, my revert did not remove a tag from a quote that had no attribution or citation. If Dream out loud and I are talking about the same thing, Dream out loud assumed that because the source was at the end of the paragraph. Per the template documentation, a citation is not needed for every sentence - every paragraph would be better. And I fixed what I feel is the only valid tag Dream out loud added, as stated on the user's talk page.[1] I also mentioned the other one.[2] Flyer22 (talk) 19:06, 17 July 2011 (UTC)

Often material from otherwise reliable sources has to be omitted. You have given one example. The OJ Simpson comparison is pov and non-WP:TOPIC. It's irrelevant whether it appeared in the NY Times. It's still unacceptable commentary which is over-the-top for an encyclopedia. We're not trying to "inflame" opinion or sell air time. We are simply trying to record the death of one person and the subsequent trial of another. Inflammatory language is best left to tabloids, not here. But we need policy reasons, as given, to specifically exclude them without the accusation of "censorship.!" Student7 (talk) 02:36, 21 July 2011 (UTC)

Student7, I disagree. The Public and media reactions section is all about the public and media's reactions, and if this case has been significantly compared to the O. J. Simpson murder trial (which it has), then this information should be included in that section. Of course it is the POV of the public and media; the whole section is about the POV of the public and media, which is perfectly acceptable. Including such information is not about inflaming anything. It is about accurately covering the aftermath. Flyer22 (talk) 11:12, 21 July 2011 (UTC)

Guys, you're in the wrong place. I recommend that you take it to WP:NPOVN. Actually, I recommend that you drop the whole thing and come back in a couple of years when nobody cares. But if you don't want to wait that long, then take it to NPOVN. There's nothing in this conversation that is actually about improving this page. Use WP:Dispute resolution, not the "re-write the guideline to support my view" game, to resolve your dispute. WhatamIdoing (talk) 04:58, 21 July 2011 (UTC)

I think my point has been missed. What I'm trying to say is that sticking 10 citations at the end of a sentence with weasel words isn't justified. For example:

Many people had positive reactions to the event.

Joe Schmoe, Jane Doe, and John Q. Public had positive reactions to the event.

A lot of content in the article is similar to the first sentence. Just because there are citations and the end does not compensate for the lack of attributions. –Dream out loud (talk) 05:55, 21 July 2011 (UTC)

WhatamIdoing, I disagree that this is off-topic and doesn't help improve this page. It does help...because it shows that there is a misunderstanding about what is laid out in WP:Weasel words and that it needs to be fixed.

Dream out loud, I still disagree with you. What you state makes no sense because you make it seem as though the word "many," "some" or "others" can never be used in relation to people and that we must always name each and every person. Or just leave it as "people" instead of as "many people," "some people" or "others." You are completely neglecting the instances when that cannot be done. Such as the fact that most women only achieve orgasm through clitoral stimulation; this is backed up by reliable source after reliable source and there are no reliable sources to dispute it. Should we just leave it as "women only achieve orgasm through clitoral stimulation"? No, because that makes it seem as though all women only achieve orgasm this way. Sources don't say "all." They say "most." And of course the sources don't specify who all these women are. And even if they did, it would be pretty idiotic to list all these women. Even more so because they are no-names.

Since it seems we will be getting no help here, I am transporting this discussion to Wikipedia talk:Manual of Style, where help is usually given to an issue that may signal that a rewrite of a guideline is needed. Flyer22 (talk) 11:12, 21 July 2011 (UTC)

I must agree with Flyer that "some" and "most" are acceptable words on Wikipedia. I would also, rather cautiously, say that most of the time, the fact that someone suspects that any given "some" or "most" might be a weasel word is reason enough to reword the sentence into something more specific. You'd be surprised how often replacing a "most scientists believe" with an "according to Dr. Firstname Lastname of Suchandsuch University" makes the article stronger. Darkfrog24 (talk) 12:01, 21 July 2011 (UTC)

In my view, although technically footnotes provide the same information as placing the information in running text, placement creates implications. If one says "some believe X^[1][2][3] the implication that the cited sources are a sample of those who hold the belief X. This is fair if it would be easy to find many other such sources, and unfair if the three sources are the only one in the world, after a thorough search, who could be found to hold the belief. Likewise, "Professor A, columnist B, and Representative C believe X" implies they are the only three sources who could be found that believe X, so if they are only a sample of many other easy-to-find sources, naming them in the running text creates an unfair implication.

It would be wonderful if there were a reliable, reasonably unbiased pollster who created polls on every conceivable topic broken out by the background of the respondents, so we could give reliable statements about any topic of interest, such as "According to Acme Polls, 78% of American scientists favor the spelling 'meter' while 61% of the English-speaking general public favors the spelling 'metre'", but no such polls exist, and we have to make do with what we can find. A balance has to be struck between remaining silent about stuff everyone knows because we can't find a poll, and making bold statements supported by a few marginal sources. Jc3s5h (talk) 12:24, 21 July 2011 (UTC)

Darkfrog24, thanks. I appreciate it. As to what you state about replacing a "most scientists believe" with an "according to Dr. Firstname Lastname of Suchandsuch University," I have to argue Jc3s5h's point on that. If he or she is "only a sample of many other easy-to-find sources, naming them in the running text creates an unfair implication." It wouldn't be fair or entirely accurate to attribute the O. J. line to only one person, when it has been compared to the O. J. trial by various people. And of course naming all these people is impossible because not all of them (which includes "regular folk") are named in the sources. Naming all of them would also be unnecessary and create that name/linkfarm I was talking about. The O. J. line above doesn't even use "most," "some" or "others." There are just a lot of times where "most," "some," "others," "majority," minority," etc. have to be used, especially if reported as exactly that through reliable sources. The orgasm example is one such case. It cannot be attributed to only one woman. It cannot be attributed to just "one researcher believes this." Researcher after researcher states "most women" or "the majority of women," although they do sometimes give an exact percent...such as 70.

In any case, the different interpretation of WP:Weasel words signals that the guideline on it needs to be tweaked a bit. Flyer22 (talk) 14:19, 21 July 2011 (UTC)

Note: Different interpretation of the guideline, regarding the issue of specificity, has also been addressed as a problem at Wikipedia talk:Manual of Style (words to watch)#"Weasel Words" - Caution or Prohibition?. Flyer22 (talk) 14:48, 21 July 2011 (UTC)

"Some" and "most" are acceptable; they are acceptable when they are either uncontroversial, or supported by an uncontested reliable source (which comes to much the same thing). Is either of these likely to be true here?

If it is, then the claim is verifiable; if no more precise claim is verifiable, follow the policy WP:V, not the guideline which implements it. That's IAR. All guidelines, including this one, are approximations to what policy would require, intended to guide, not command. Septentrionalis PMAnderson 19:54, 21 July 2011 (UTC)

Thank you, Pmanderson. And, yes, the sources are used in ways that are truthful and accurate to what they are backing.

Would either of you be willing to tweak WP:Weasel words to keep these kinds of misunderstandings of the guideline from happening? Flyer22 (talk) 21:38, 21 July 2011 (UTC)

If I did, it would become rather less supportive of your position than it is now. As it is, it only requires attribution, which is not enough. Septentrionalis PMAnderson 02:01, 22 July 2011 (UTC)

Why would it need to be less supportive, given what is stated above about the problem with specificity? In addition to what I stated about name/linkfarms, Jc3s5h made really good points as well. You said attribution is enough when the line it is backing is uncontroversial and/or when used accurately, but you are also saying that the guideline should speak of more than just attribution? On that point, the words "clearly attributed" in the guideline are what started this dispute. It doesn't just say "attributed." As seen above, Dream out loud and I interpreted "clearly attributed" differently. Even if your rewording of the guideline were to lean even more towards specificity (which isn't too different than what it already does), I would not have a problem with it...just as long as it is clear that "some" and "most" are acceptable in cases similar to the ones above. Or maybe "clearly attributed" needs to be changed to "attributed." I don't know, but when there are these kinds of different interpretations of what should be a simple guideline, it is clear to me that the guideline needs to be clearer in what it means. People are tagging properly attributed sentences with the "whom" template and such, when as WhatamIdoing says, "...the documentation for that template. It has apparently restricted its use to unsourced statements since the template's creation." Flyer22 (talk) 13:56, 22 July 2011 (UTC)

Isn't there pretty much always a problem when a single sentence is supported by 7 citations? That seems like a serious red flag to me. john k (talk) 15:43, 22 July 2011 (UTC)

Not when it's covering various opinions, as is the case here. The line mentioned above says: The trial has been compared to the O. J. Simpson murder case, both for its widespread media attention and reported "shock" at the "Not Guilty" verdict. Well... Since many have compared it to the O. J. trial for the reasons mentioned in that line, it is more accurate to not attribute the line to one person...such as saying "According to John Cloud, [etc., etc.]" As Jc3s5h stated above: "Professor A, columnist B, and Representative C believe X" implies they are the only three sources who could be found that believe X, so if they are only a sample of many other easy-to-find sources, naming them in the running text creates an unfair implication. More than that, I often like to be thorough in my sourcing. One or two sources could suffice for that line without mentioning any name, but, like I stated, I am often thorough (whether I do multiple ref stacking or not). Flyer22 (talk) 17:55, 22 July 2011 (UTC)

I think this is very much the wrong way to look at it. Isn't combining multiple sources to make a judgment found in none of those original sources original synthesis? john k (talk) 05:41, 23 July 2011 (UTC)

It's not WP:SYNTHESIS because I'm not combining "material from multiple sources to reach or imply a conclusion not explicitly stated by any of the sources." It is explicitly stated in the sources that the trial has been compared to the O. J. Simpson trial and why that is -- media attention and shock/outrage at the verdict. You can check the sources yourself. And if you have, you must have read them wrong. Any source attributed to that line not comparing the trial to the O. J. trial can be removed, but I doubt that any of them are not. I read through all of them before adding them. WP:SYNTHESIS is quite clear that it would only be synthesis if the published material is being used "to advance a new position, which is original research." There is no "new position" in relaying that the trial has been compared to the O. J. Simpson trial and why; this is stated explicitly in the reliable sources, with some people calling it "O.J. Number 2." As WP:SYNTHESIS states, "A and B, therefore C" is acceptable only if a reliable source has published the same argument in relation to the topic of the article. And that's what I've done. If I had created WP:SYNTHESIS, then that would have been stated above by the other editors, and they would not have agreed that what I have done is acceptable. I get that you don't like ref stacking involving multiple references, and you'd probably even prefer never to use "some" or "most" on Wikipedia (as would I in a lot of cases), but that is your preference. It has no bearing on whether or not my using "most," "some" or "others" is acceptable, given the reasons stated above, which, according to a few editors of this project above, it is. And as also shown above, the O. J. line in question does not use "most," some" or "others." Flyer22 (talk) 12:44, 23 July 2011 (UTC)

I'm happy to use "most" or "some" in wikipedia articles, if the sources support it. But I should be able to have one citation to back up that claim. The more citations, the weaker the evidence, it seems to me - else why do you need seven citations for a single claim? You have a bunch of sources that individually compare the Anthony case to the Simpson trial. Okay, that's fine. But why do we think this is, in itself, a significant piece of information to include in that article? And why do we need seven citations? Do none of these sources say "the case has been compared to the OJ Simpson trial"? Because that's what we need. If we have that, then we don't need all the other sources. If we don't have that, bringing in specific comparisons to the OJ trial is problematic because it is unclear, in the absence of a source, what the significance of these comparisons is. How do we judge that this, as opposed to other comparisons that have surely been made, is worthy of inclusion in the article? john k (talk) 21:06, 25 July 2011 (UTC)

Like I stated, "I often like to be thorough in my sourcing. One or two sources could suffice for that line without mentioning any name, but [I didn't prefer that in this case]. I get that you don't like ref stacking involving multiple references... ...but that is your preference." So I'm not sure why you're arguing preferences. Saying "But I should be able to have one citation to back up that claim. The more citations, the weaker the evidence, it seems to me" is your opinion. That's not even how Wikipedia usually works -- using one citation for one line. And it is your preference not to use several citations for one line. I, however, like to use more than one, and sometimes several. One reason I did so in this case is because of what I stated above: Since many have compared it to the O. J. trial... ...it is more accurate to not attribute the line to one person...such as saying "According to John Cloud, [etc., etc.]." The other reason I ref-stacked? Well, when I use three or more, it is usually because I am using those sources for other things in the article as well, and I feel that it is a waste not to attribute those same sources to other lines they back up. That's often how I arrive at stacking three or more sources to one line. It's rarely ever because a line is contested. As for including the O. J. aspect? Well, mentioning that the Casey Anthony case has been compared to the O. J. Simpson trial is a significant piece of information because it has been significantly compared to that case, especially regarding the shock/outrage at the verdict. So why should this information be excluded from a section discussing the Public and media reactions to the case, when the public and media reactions to the case have largely been about comparing this trial to the O. J. Simpson murder trial? That is the better question. And, no, we don't need a source specifically saying "the case has been compared to the O. J. Simpson trial." All we need are sources showing it being compared to the O. J. Simpson trial. It is not then original research or, more specifically, WP:SYNTHESIS to say that it has been compared to the O. J. Simpson trial when the reliable sources are quite clearly comparing the two. But, yes, some of sources I included specifically state "Inevitably, the Anthony trial will draw comparisons to the OJ Simpson trial." and "Casey Anthony: Echoes of OJ Simpson For Veteran Trial Watchers," etc. etc. And the section does further specify how Anthony's case compares to the O. J. Simpson case, such as in terms of ratings.

The bottomline is we are disagreeing more than we are agreeing, and I didn't bring this discussion here to get into a debate about personal preference regarding ref stacking. I brought it here to get clarity on whether or not "most," "some" and "others" are perfectly valid words to use on Wikipedia when attributed to people, and I got that. We disagree on my sometimes-referencing style. There is no reason to keep debating it, since my mind is not going to change and it seems yours is not either. Flyer22 (talk) 22:33, 25 July 2011 (UTC)

Trial unprotection

Okay, I am going to unprotect this page. Any further acrimony on the MOS page will be taken very dimly. Please do not edit war or tag it. Let's move on. Casliber (talk · contribs) 05:33, 24 July 2011 (UTC)

Shame, Casliber. No tag-bombing, when one fails to muster consensus support for a view? You are clearly insensitive to the reason that a certain editor or two comes here in the first place.  ;)--Epeefleche (talk) 17:28, 25 July 2011 (UTC)

This decision redefines "consensus" to be the view of (at best) a thin, divided and vocal minority - as the text of this page always has been.

• We began this long discussion with Tony's proposa1, which would very grudgingly have permitted both dashes and hyphens. Support and oppose divided almost equally; more importantly, most of the opposition thought that it was too stringent, and should have permitted hyphens more readily. That's one majority Casliber's close ignored.
• We ended it with this poll. Read point 5b; a majority "prefers" dashes in that context; but the strongest posssible minority (14 out of 29, at least) explicitly wishes to make them optional. Casliber's close also ignores this large block of sentiment.

And this is called "consensus." What civil comment is possible? Septentrionalis PMAnderson 20:09, 25 July 2011 (UTC)

Whether to make hyphens optional in the rare (nonexistent?) cases where it is undesirable to rephrase a prefix to a multi-word phrase doesn't seem like that big of a deal, to say the least. You're making it sound like there was an important issue under debate. I'll note that Noetica, the drafter here, didn't even seem to like this item, yet wrote it in the draft; it seems not all of the dissenters share your evaluation of what happened here. You can always try to get it changed, or insert some language about making this "optional" if you like. ErikHaugen (talk | contribs) 20:58, 25 July 2011 (UTC)

"Weasel Words" - Caution or Prohibition?

As noted by Flyer22 above, I had posted the following question to Manual of Style (words to watch) but have not, as yet, received a response to my query (save for editors involved in the discussion that generated the question).

I am reposting it here (in a more generic fashion) to, hopefully, better facilitate an informed response.

Some editors contend that WP:WEASEL is prohibitive and that non-specific wording, where challenged, can be legitimately struck unless rectified by the addition of language to increase specificity. My contention is that non-specific language, when the language is consensus developed and acceptable, can be perfectly legitimate under WP:WEASEL.

Comments from editors particularly versed in this subject are kindly solicited and appreciated. JakeInJoisey (talk) 11:41, 24 July 2011 (UTC)

(reposted, this was also posted here)I think JakeInJoisey actually made up the claim that other editors are taking WTA as prohibitive. JakeInJoisey said on the article talk page "IMHO, your interpretation of WP:WEASEL as some mandate for specificity is erroneous," which no one had said, certainly not myself. He also [said in the original posting], "My contention is that non-specific language, when the language is consensus developed and acceptable, can be perfectly legitimate under WP:WEASEL" and again no one is contending otherwise: rather, when non-specific language is actually challenged, and therefore not in accordance with full consensus, we need to look closely. My contention would be that per WP:BURDEN, non-specific language needs sourcing if challenged. WP:WEASEL is just a way of helping editors notice when we may need better sourcing or better adherence to sources, as well as a call to good writing. In this specific case, it has been demonstrated that the material in question can be well-written without weasel wording. BE——Critical_{__Talk} 14:25, 24 July 2011 (UTC)

...when non-specific language is actually challenged, and therefore not in accordance with full consensus...

That's an erroneous interpretation of what constitutes "consensus" (which needn't be and rarely is universal), mis-represents the thus far consensus-based and editor-commended evolution of language used in that article, ignores an ongoing, active attempt to reach consensus on numerous related issues and is, rather clearly, a transparent attempt to inappropriately disrupt my solicitation for informed opinion on the MOS question raised which is independent of specific article issues. It would be appropriate, IMHO, for any admin to delete your disruptive and irrelevant observations (and my responses) from this generic inquiry and I urge any administrator to please consider doing just that.JakeInJoisey (talk) 15:07, 24 July 2011 (UTC)

The 'bright line' is WP:V. You cannot make a consensus agreement to override verifiability. The problem with most language that fails WP:WEASEL is that it cannot be verified. "Many people say that The Scrotums are the greatest band of 1977" begs the question "OK, who exactly says this". It may be in some cases that one has reliable sources that themselves say "Most reviewers agree that Escape to Tordmorden was his worst film", in which case the apparently less specific language can itself be sourced. In some cases, one can have the majority of mainstream sources agreeing that "1988 was a terrible year for vegetable marrows", but a minority opinion that one would want to reflect, so "It was widely held that 1988 was a terrible year (source), except on the Isle of Wight (source). But you cannot agree by consensus to language that you cannot verify - "some say he was born in a cave in the Andes. All we know is...he's called the Stig". Elen of the Roads (talk) 17:26, 24 July 2011 (UTC)

LOL "The Scrotums," and exactly, thank you for responding. I know people don't like to touch this santorum thing. BE——Critical_{__Talk} 18:03, 24 July 2011 (UTC)

Thanks for your response, but I'm not sure you've addressed my question as I attempted to present it. However, I didn't explicitly address WP:V because, in my presentation, I was assuming that WP:V would be considered a given. Of course as you stated and beyond question, any "Weasel Wording" failing WP:V could be summarily dismissed.

Given my lack of clarity, please allow me to rephrase the question...

Some editors contend that WP:WEASEL is prohibitive and that non-specific wording, where challenged, can be legitimately struck unless rectified by the addition of language to increase specificity. My contention is that non-specific language, when the language is WP:V, WP:RS supportable by citation, consensus-developed and consensus-acceptable, can be perfectly legitimate under WP:WEASEL.

Does that change your perspective and understanding as to the question I'm posing? Thanks. JakeInJoisey (talk) 18:16, 24 July 2011 (UTC)

comment Campaign for "santorum" neologism is a quagmire of nitpicking. There are those that believe the article should stay, but are determined that it should not reflect badly on the Savage. Then there are those that believe that the article should really go or be merged, but if it is going to stay then it shouldn't aid the campaign against Santorum. As a result it ums and ahs, as it stutters from clause to clause. John lilburne (talk) 21:09, 24 July 2011 (UTC)

Of those two, I'd go with "caution." I don't want anyone to use this as an excuse to ignore legitimate complaints about weasel words, though. I stand by what I said earlier: The fact that someone questions whether the words are weasel is a very good reason to render the text more specific (but people will abuse this if we make it into a rule). Darkfrog24 (talk) 22:05, 24 July 2011 (UTC)

Darkfrog, please keep in mind 2 VERY relevant aspects assumed in my question. I'm referencing use of a non-specific that is 1. WP:V and 2. clearly consensus approved...and I'm not sure that a rule change is warranted either, more of a clarification. The phrase "can deny" in WP:WEASEL is, imho, spot on and suggests a plausibility for consideration NOT a conclusion and consequent prohibition. Thanks. JakeInJoisey (talk) 22:18, 24 July 2011 (UTC)

Well, that's theoretical, and in such a case I would agree that the weasel words would be perfectly acceptable. But where they've been and are being challenged, and aren't necessary to writing the material, that's different. That's the case which prompted this thread. So we're in danger of providing a license to stonewall in the name of consensus (not accusing anyone). BE——Critical_{__Talk} 02:04, 25 July 2011 (UTC)

Are we actually talking about WP:WEASEL here, or something else. If "it is a widely held view that he is a tv phenomenon" actually sources to "Richard Hammond says he is a tv phenomenom", then maybe WP:UNDUE is what you are looking for. If words are questioned, sources should clarify, and the language should match the source. Elen of the Roads (talk) 13:30, 25 July 2011 (UTC)

The actual situation? These two diffs, and of course one of them is my own version, but at least that shows it's possible to do without weasels: [3] [4]. JakeInJoisey might have more information for you, but this is my take. Justifying weasels when you don't need to write with them is... to be avoided. BE——Critical_{__Talk} 14:00, 25 July 2011 (UTC)

While I would be delighted to further engage Becritical's position in this space, I'll defer to the purpose of this talk page. If any editor is interested in further discussion as to the specifics of how this issue relates to the specific article mentioned, your input would be MOST welcome in the article talk page. Thanks. JakeInJoisey (talk) 14:19, 25 July 2011 (UTC)

I was responding to a request for information. BE——Critical_{__Talk} 16:08, 25 July 2011 (UTC)

• Note - Editor opinion is also solicited on the question of current template language in Template talk. JakeInJoisey (talk) 16:18, 25 July 2011 (UTC)

Prohibit, permit, prefer--what do the WT:MoS regulars and visitors think?

I have a bit of a theoretical question to run by this crowd. EDIT: Is this how we should be interpreting the following three terms?

When two mutually exclusive styles are permitted in the MoS, it means that no editor should change any article from one style to the other without first establishing or at least gauging consensus for the change.

When two styles are permitted but one is preferred, any editor may change the article to the preferred style without first actively establishing consensus. Consensus is assumed unless the change is challenged.

When a given style is prohibited, any editor may change the article from the prohibited style without first actively establishing consensus. Consensus is assumed unless the change is challenged—legitimate challenges will be rare.

Darkfrog24 (talk) 22:13, 24 July 2011 (UTC)

What's the question? As a statement of fact it's probably also true that any editor may change things from preferred to prohibited without seeking consensus; I see it almost every day. Are you drafting new guidelines here to try to reduce the prevalence of such? Dicklyon (talk) 22:41, 24 July 2011 (UTC)

I want to be able to answer the question "What does it mean to change the MoS?" I am not in the process of drafting a new guideline. Feel free to blow this off. Darkfrog24 (talk) 01:11, 25 July 2011 (UTC)

No, this is not a legislation, outside of certain disruptive editors' imagination. It's guidance for the perplexed; the audience of a useful MOS, like other guidelines, would be editors who want advice, not munchkins who are "editing" articles they do not understand to conform to guidance they do not understand.

• If both are permitted, use either; usually the one the article already uses, if the point has come up before.
• If one is preferred, use it if you are uncertain what to use; if you are certain what to use, and it is not what MOS prefers, think twice whether you have a clear reason to do otherwise. A useful guideline would give a reason for a preferences (and a universally convincing reason for mandates).
• If we have a mandate, there should be absolutely no question that all legitimate objections have been considered and preferably assuaged. Septentrionalis PMAnderson 23:18, 24 July 2011 (UTC)

• Somewhat tricky to generalise - and best taken on a case by case basis. We shouldn't have conflicting MOS rules and if there are, we should try and resolve it beforehand. Casliber (talk · contribs) 00:14, 25 July 2011 (UTC)

• But this isn't about conflicting MOS rules (there are many and I often harp on the need to harmonize them). It is, in part, about MOS permitting two different styles, such as either American English or British English is permitted.

• For Septentrionalis: "the audience of a useful MOS, like other guidelines, would be editors who want advice, not munchkins who are "editing" articles they do not understand to conform to guidance they do not understand." Let's see how that concept would work out in practice. Suppose said munchkin encounters the Quadratic equation article, which some vandal has renamed "Quadratic Equation". The munchkin has flunked algebra and doesn't understand the article, and he doesn't understand why Wikipedia articles use sentence case titles (I was taught title case in school). But being an unusually anal munchkin, he is for some obscure reason aware of the WP:CAPS guideline. He uses Google Books to confirm it isn't considered a proper noun. So why isn't that enough for him to know before changing it? Why does it matter if he can distinguish the x² terms from the x terms, before he can uncapitalize "Equation"? And how would we deal with 10,079 articles without munchkins? And if we can't, then why does the rest of the page matter? Art LaPella (talk) 00:24, 25 July 2011 (UTC)

I agree with Art that it is okay to have some people who are experts on subject matter and others who are experts on style. We can't expect everyone to be great at both. That's why the world has professional proofreaders (for now). Darkfrog24 (talk) 01:14, 25 July 2011 (UTC)

On the other hand, as I already said, “I don't think an article with major WP:NPOV/WP:OR/WP:V problems but a professional-looking style would be better than the same article with a crappy style. If anything, the latter is less likely to deceive readers. (This is why I don't usually copy-edit articles unless I have at least a vague idea of what they're talking about and know that what they say is at least vaguely plausible.)” Anyway, if an article has been stable at a given title and it's moved without a valid rationale, moving it back is OK in any event, regardless of whether the MOS happens to mention that particular point of style or not. (Also, bots, which many people here are so fond of, are unable to do such things as “us[ing] Google Books to confirm it isn't considered a proper noun” or whatever.) ― A. di M.​_plé​^dréachtaí 10:59, 25 July 2011 (UTC)

If you "don't usually copy-edit articles unless I have at least a vague idea of what they're talking about", and presumably the munchkin shouldn't uncapitalize "Equation" because he doesn't know algebra, then it isn't enough to argue more than that a bad article with good style is worse than the same bad article with bad style. You have to argue that if there is an x% chance that an article you don't understand might be a bad one, then the harm that comes from the x% chance of copyediting a bad article, exceeds the good that comes from the 100−x% chance of copyediting a good one. I believe that hypothesis leads to absurd, Swiftian conclusions.

The proportion of bad articles among those that you or the munchkin does understand is probably x% or more, because it's more fun to write a joke article about something simple like the president's butt than about something arcane like quantum mechanics, and similarly for careless articles. Therefore, according to the hypothesis, if a person doesn't make any effort to evaluate the rest of the article he copyedits, it would be better if he didn't copyedit at all. That description fits AWB, because although someone correcting all instances of "occuring" to "occurring" is expected to avoid a context like a list of misspellings, he is not expected to evaluate the quality of each article, and if he were so expected then there would be no reason to use AWB at all. So if the harm done by the chance of improving a bad article overrides the benefit of the chance of improving a good article, there should be no AWB. Actually, there should be, mathematically speaking, but only to introduce deliberate errors throughout Wikipedia! The benefit of discrediting the bad articles exceeds the harm of ruining the good ones. Reductio ad absurdum. Art LaPella (talk) 15:51, 25 July 2011 (UTC)

You have picked an example which is unusually straightforward; the question about quadratic equation is not what the sentence case capitalization of "quadratic equation" is; but whether to use sentence case for our titles at all. (This is one of the very few issues where this page prohibits something which I have not seen seriously disputed, but I may have missed it.) Once that decision is recognized, the munchkin need not come here for a ruling at all; his sources will say In solving quadratic equations, not (since the early Victorians) In solving Quadratic Equations.

But for my part, I would rather not have somebody who doesn't understand the article copy-editing it at all; the chance of his removing or inserting a ± in the wrong place are far too great. If the function of this MOS were to give instructions contrived by ignorant pedants to be blindly followed by semi-literates - as is all too frequently supposed - we should indeed be better off without any; but I do not believe this. Septentrionalis PMAnderson 16:44, 25 July 2011 (UTC)

My example was an article title "Quadratic equation", not the article text "in solving quadratic equations". My own Google Books search above shows that it's usually "Quadratic Equation" or even "QUADRATIC EQUATION" in a title, so if we "follow sources" I suppose we would use one of those choices, perhaps randomly using MOS:ALLCAPS for half our article titles.

As for your second paragraph, it seems inconsistent with the article quality you can see by clicking "Random article". Much and perhaps most of the Manual of Style was unfamiliar to me before I encountered it, and yet even then my copyediting was much better than what random articles were like (especially at that time). I reasonably concluded that Wikipedia doesn't take the Manual of Style seriously, and it wasn't until years later that I took more of an interest in it.Art LaPella (talk) 17:05, 25 July 2011 (UTC)

I persist in the irresponsible suggestion that Wikipedia would take MOS seriously, if it were less WP:FRINGE. ;-> Let me know if there is ever a movement to try this. Septentrionalis PMAnderson 19:20, 25 July 2011 (UTC)

If there's a silent majority we're flouting, we now know it takes more than an RfC to find it. Art LaPella (talk) 20:02, 25 July 2011 (UTC)

But it doesn't take even that much to find a substantial minority we are ignoring; the section on #logical quotation shows that just as well. Septentrionalis PMAnderson 20:34, 25 July 2011 (UTC)

Titles of whole books are normally in title case, and often titles of chapters are in title case as well, but in my experience titles of encyclopaedia entries seldom are. ― A. di M.​_plé​^dréachtaí 22:18, 25 July 2011 (UTC)

Well, as for me, psychologically there's also the “I'd rather have nothing to do with a hoax/BLP violation” issue. YMMV. ― A. di M.​_plé​^dréachtaí 22:18, 25 July 2011 (UTC)

• On first blush, DarkFrog’s interpretation of how editors should take MOS teachings to the field makes sense to me. Greg L (talk) 01:04, 25 July 2011 (UTC)

Yes, it sounds sensible, which is why I wondered what the question was. On closer inspection, however, one has to wonder, since the MOS (or the main MOS page at least) doesn't mention anything as either "permitted" or "prohibited". So I'm wondering if he's thinking of changing it to use such language. I bet that would be fun! Dicklyon (talk) 04:51, 25 July 2011 (UTC)

"Permitted" isn't in the MOS vocabulary for the simple reason that, as some love to point out, it is only a guideline, implying that it does not have to be followed. ;-) --Ohconfucius ^¡digame! 06:36, 25 July 2011 (UTC)

It might not use the words, Dicklyon, but it does perform the actions: Both using and omitting the serial comma are permitted. British English spelling is preferred on articles with ties to British subjects. American punctuation is prohibited.

I wasn't thinking about changing the language of the MoS, but if you think there's any problem that doing so would solve, hey, I'm in. Darkfrog24 (talk) 12:52, 25 July 2011 (UTC)

Much through I hesitate to involve Darkfrog in my heresies, I cannot but agree; if we stopped playing ostrich, admitted the existence of American punctuation, and put the conflicting arguments ("American punctuation may mislead some readers about trailing commas; British punctuation is very difficult to proof-read and get right"), that issue would go away. Septentrionalis PMAnderson 19:20, 25 July 2011 (UTC)

Proposal to standardize use of em dash instead of en dash for full date range

For date ranges that include only years, unspaced en dashes work better (i.e. 1995–2002). But for date ranges that include the month and/or exact days of the range (prime example being exact DOB–DOD ranges), I believe a spaced em dash fits better than the current standard, a spaced en dash. It seems to make clearer the distinction between one whole date and the other, and it makes each date seem more independently stated. Spaced en dashes don't accomplish that very well in these types of ranges. See, for example:

En dash (current standard):
Amy Jade Winehouse (14 September 1983 – 23 July 2011) was an English singer-songwriter.

Em dash (proposed standard):
Amy Jade Winehouse (14 September 1983 — 23 July 2011) was an English singer-songwriter.

Thoughts on this?. — CIS ^(talk | stalk) 10:39, 25 July 2011 (UTC)

On my computer, it's just a bit too jarring. Better to avoid the dilemma with the little word "to"

Amy Jade Winehouse (14 September 1983 to 23 July 2011) was an English singer-songwriter.
Abraham Lincoln (February 12, 1809 to April 15, 1865) was the sixteenth President of the United States of America.

—— Shakescene (talk) 10:50, 25 July 2011 (UTC)

That's ever weirder; I can't recall ever seeing a bio in English doing that. Even (b. 14 September 1983, d. 23 July 2011) would be better. (And don't let me started again on whether days and months, rather than just years, are always so vital as to deserve being in the first sentence.) ― A. di M.​_plé​^dréachtaí 11:04, 25 July 2011 (UTC)

Hmm, I don't know... fullstops seem to be becoming increasingly deprecated in modern English, and on Wikipedia in particular. (Both of) your proposed format(s) seem to mesh the dates a bit from my view of things; I think the spaced em dash works best to make the distinction between the two dates as clear as possible. And as for why the full DOB/DOD range should be in the opening sentence, I'd say it helps to give some body to the intro., especially in articles that are lacking content. Plus it's only a few extra words, I don't see that as much of a problem. It looks nice as well; it's Wikipedia's signature stylistic feature in biographical articles. — CIS ^(talk | stalk) 11:14, 25 July 2011 (UTC)

• Oppose. This suggestion seems to come from nowhere; it is not among the ideas the came up in the en dash discussions, and no support from style guides has been hinted at. The current scheme of spaced en dashes is widespread and widely accepted; is anyone objecting to it? I see no reason to fish around for anlternatives when something is working this well. Dicklyon (talk) 14:53, 25 July 2011 (UTC)

• Oppose as well. En-dashes are already standardized for dates. This change would only complicate and muddle a relatively straightforward and settled usage. —Torchiest ^talk_edits 15:28, 25 July 2011 (UTC)

• Oppose. I’ve never seen this anywhere and cannot find support for it in any style guide. It’s one thing to discuss which well established usages to allow, but quite another to invent new ones. If we do the latter, I don’t see how anyone else will understand what we’re doing. I’m clearly no fan of spaced en dashes, but in this case there’s no ambiguity (I don’t normally subtract dates), and use in Wikipedia, especially in bios, is well established. JeffConrad (talk) 21:50, 25 July 2011 (UTC)

• Oppose as well, for the reasons cited above by Dicklyon, Torchiest, and JeffConrad. And as Shakescene says, it's a bit too jarring visually—there's just too much space between the 1st date's year and the 2nd date's day, with no evident gain in readability or clarity. (A case might possibly be made for an unspaced em dash, which would only create roughly the same horizontal distance between the dates as a spaced en dash. But again, why? The spaced en dash seems well established, and there's no compelling argument for changing it. Indeed, there's no argument at all presented in to support the proposal.) Best leave well enough alone. Jack --Jackftwist (talk) 21:45, 27 July 2011 (UTC)

Is a slogan a quote

I'm an outside observer to an edit war, based on the MOS. What I will call unoriginal TV stations use the same slogans in different markets in order to position their station image. I make this a distinction from an original or unique quote, its merely a marketing identity. So some other editors have created articles detailing the mass usage of these common slogans. And at least one IP editor is going around wikilinking the usage of the slogan to the articles. And another editor is spending a lot of edits deleting the links because he claims they are quotes. How can we draw the distinction? I say it is in the lack of originality itself, if there is no uniqueness to the phrase, its not a direct quote. And the mass usage of that unoriginal phrase is documented in the wikilinked articles this editor is delinking. Trackinfo (talk) 05:17, 26 July 2011 (UTC)

This one could depend on how it's being used. Can you give us an example or two of these slogans? Darkfrog24 (talk) 06:19, 26 July 2011 (UTC)

Eyewitness News and News Channel are the two I initially noticed. I see there is also Action News and NewsWatch. Trackinfo (talk) 15:46, 26 July 2011 (UTC)

US copyright law does not allow common expressions to be trademarked. For instance, a store chain can not trademark “Cash & Carry”. I assume the same goes for “Eyewitness News”, “5:00 News”, etc. Perhaps this lack of uniqueness is a factor in whether one needs to put these in quotes. Like Darkfrog24 commented, it helps to have examples. The examples you provided are out of context; which is to say, there is no context. Can you provide some example sentences? Greg L (talk) 15:58, 26 July 2011 (UTC)

The article I became aware of this edit war was KCOY-TV#Newscast Titles Trackinfo (talk) 17:32, 26 July 2011 (UTC)

Nothing in here jumps out at me as being inappropriately linked. Has the link that you noticed been removed? Darkfrog24 (talk) 18:04, 26 July 2011 (UTC)

Apologies, but I believe my position is being misrepresented. My point has nothing to do with copyright or trademark. My point is that, as you can see in KCOY-TV#Newscast titles, inside 2 of the show titles, 2 individual phrases are linked. Note that these are not links to an article about the show, but rather articles that describe the fact that a wide number of different US tv news stations use similar branding. To me, this is essentially the same as if in an article someone linked The Big Bang Theory as The Big Bang Theory. I argued that the linking of individual words or phrases within a title is equivalent to link individual words or phrases within a quotation, which is not allowed per the MOS. Trackinfo mentions that this is a "marketing identity", but please not that when a tv station uses the phrase Action News in their title, that isn't to link it to some larger, national branding for the news station; its simply that it is a common name used in many stations. Qwyrxian (talk) 00:17, 27 July 2011 (UTC)

Yeah, I find it quite irritating when I follow a link in a track list and instead of an article about the song I get to one about the thing the song is named after. ― A. di M.​_plé​^dréachtaí 17:59, 27 July 2011 (UTC)

In this case, I think the commonality of the use of the phrase, at the exception of the superior article about this station's specific use of the phrase, is better than no link at all. In the case of Eyewitness News, at least for a period of time, it was the domain of, I'm learning as I go Al Primo who created Happy talk. A specific marketing company selling the phrase to be used (and later copied) . . . This is the value of wikipedia and wikilinking. I knew of the marketing of the Eyewitness News name, but someone has made an article that puts things in perspective and names the names. I would have never found it on wikipedia without finding these links. Trackinfo (talk) 20:44, 27 July 2011 (UTC)

Except that if we keep reading Eyewitness News, we see that it wasn't primarily a marketing company selling; it was "copied by many other stations" and "A separate, but mostly unrelated, Eyewitness News format was developed....." In other words, the various stations that use Eyewitness share nothing in common except for the name itself. Some share a certain format...but, let's face it, most local news broadcasts share a lot of similarities. To me, linking them borders on being deceitful, since it implies a connection to other tv station news shows, when in fact, there may not be, or, if there is, exactly how it is linked is completely unclear. I think A.di.M. gives a good comparison with song titles, and I would argue that such linking is also inappropriate. Qwyrxian (talk) 22:21, 27 July 2011 (UTC)

I disagree with the analogy of words within song titles. While the lack of originality of the phrases in their wider usage is simply copying what is un-copyrightable (the slogan), the phenomenon specifically of Eyewitness News does have a pattern in the major o and o markets. Once guy, the aforementioned Al Primo started it on the flagship station for Westinghouse Broadcasting, then it spread to all other Westinghouse stations, then he got a job at ABC, the exact same slogan went to their flagship station then to all other ABC o and o's (except in markets where Westinghouse had the slogan first--the law of public perception of copyright would apply). After it was successful in the majors, lots of stations in the minors copied it. The fact that this slogan was copied is a phenomenon, a worthy subject for wikipedia to cover and internally link. Do our voluntary job as wikipedia editors, complete that chain of information. Bill Drake didn't sell boss radio to everybody either but every market had something like it. Successful media concepts are rapidly copied in local markets because there are so few people with original ideas in media. Back to the main point: these formats are significant, related stories to the slogan, not merely something named after something else. The articles connected do one of the functions wikipedia does so well; it correlates information. A list like these may not exist on any other single source, but the efforts of one motivated wikipedia editor and subsequent contributions congeals the information. Thats all good, thats not misleading. I mention on my user page, one of the best things that has happened to me here on WP was when I started a relatively short article about Captain Mikey the creator of Album-oriented rock, the AOR format. Another editor found the link, did more research, made huge contributions and turned it into a 72 source monstrosity. Things like that don't happen when you hide links to supporting articles. On the other hand, deliberately deleting links, hiding the information because you alone think the destination is unimportant, is a bad imposition of POV. Trackinfo (talk) 02:32, 28 July 2011 (UTC)

Dashes: a new draft

Having considered the voting and the long discussion (see some details above), I offer a new draft for consideration at this central forum for MOS, now that business appears almost complete at the dedicated subpages. Noetica^Tea? 03:05, 13 July 2011 (UTC)

Suggested procedure

1. It is essential that our concluding discussion be kept orderly, focused, and respectful. With this in mind I am suggesting a framework. I urge editors to use this framework, or alternatively to propose rational variations of it. I urge all editors to work to keep the discussion easy to follow, and to refactor if that will help.
2. I request that no changes be made to the draft during discussion of it. Please instead propose new versions for parts of it, with localised discussion. Again, it will be easy for the process to disintegrate if we are not vigilant to stay on track.
3. I propose that we rule out certain distractions especially, and that editors join in removing such distractions to a special subsection I have set up to receive them:

   • personal attacks or innuendos

   • threats, or expressions of intent to take admin or ArbCom action against any editor

   • political statements and manifestos

   • unnecessary reference to past conflicts

   • insistence that the Manual of Style be something other than just that (its status is definitely not our present topic)

   • reflections on the role of MOS generally within Wikipedia

   • other obvious irrelevancies or contention

4. I suggest that there be no voting at this stage; let's see how things take shape.
5. If any editor wants to propose a rewritten whole draft, please bear in mind the consequences. Discussion should have one overall focus at a time, and we can all end up confused and directionless. Therefore: Just as I gave 24 hours' notice before posting a whole draft, please give similar notice of any further whole draft.
6. The single question we face is this: What dash guidelines will best reflect the clear results of the voting?

Notes on the draft

1. The draft does not reflect my own preferences as its proposer; it will not satisfy any individual's preferences fully, nor should it aim to.
2. As a practical necessity, I have assumed that consensus means clear majority assent.
3. Some preferences in the voting are for more than one recommendation (for en dash or hyphen, in a given situation). This is heeded where it is a majority preference; but if the majority accepts a guideline simply as given, that simple endorsement has been respected as best reflecting consensus.
4. Some matters have been set aside (partly or wholly) as better treated in general sections of WP:MOS, since they apply beyond the guidelines we deal with here. They include:

   • the desirability of recasting (needs a new section of its own)

   • the fact that exceptions are sometimes in order (already thoroughly covered)

   • use of hard spaces (already covered in WP:MOS, but it may need adjusting)

5. The draft is much longer than the existing section. It has to be, since this, along with hyphens, is one of the most difficult areas in modern punctuation. Editors will suggest ways to shorten it; but the voting has shown how easily brevity leads to entirely wrong interpretations. We need to guard also against intentional misconstrual.
6. Cases like a pro-establishment–anti-intellectual alliance are not treated with their own special principle. I found that they were covered by another more basic guideline that is included. This requires discussion. Can anyone think of a case that warrants, in the MOS context, a separate provision for such cases?
7. As an aside, we need to revisit the colors used in the {xt} and {!xt} templates (for readers unable to distinguish red and green). Meanwhile, I have preceded all red examples (of what not to do) with an asterisk:

   *−10–10; −10 to 10

The draft itself (click on the "show" link in the green bar)

Discussion of the draft

[Please: orderly, not interrupting, clearly signed]

At first reading, this looks pretty good. I still think we need to mention the use of the en dash (or em dash??) as a separator when it's neither in a sentence nor in a list, as in the Harry Potter and the Deathly Hallows – Part 1 case we discussed – this seems to come up from time to time, in article titles at least, and people usually seem to opt for a spaced en dash (or some alternative other than a horizontal line) when the matter is considered.--Kotniski (talk) 04:07, 13 July 2011 (UTC)

Thanks, Kotniski. And thanks for breaking the ice with your draft, which influences the one I offer here. I agree that more work needs to be does on the point you mention. Luckily, it is self-contained and probably not a major cause of difficulty. Why not propose a neat modification to cover it? Noetica^Tea? 06:23, 13 July 2011 (UTC)

Noetica, I've looked only cursorily through your draft, and I must say that it's impressive. (Kotniski's was impressive too, but I believe this more recent draft takes the ball and runs with his improvements.) I was slightly concerned at the length of Noetica's draft, and we need to scrutinise it for "trimming opportunities"; but then, putting myself in WP editors' shoes, I learn something from every example, and examples are the key to conveying professional-standard typography. When this is tweaked, the final version might well be worth printing out and distributing to my clients. I'm most interested in the opinions of other editors. Thank you indeed, Noetica. Tony (talk) 06:43, 13 July 2011 (UTC)

I think it looks pretty good, but a few comments:

• The order of the recommended/not recommended practice in the examples should match that in the text that precedes it, rather than the opposite, as in

  Use words, not en dashes, if negative values are involved. An en dash might be confusing.

  *−10–10; −10 to 10

• Spacing of range dashes where one or more elements contains a space: either spaced or unspaced use should be allowed. The polling seems to indicate considerable support for allowing both, and the support for closed-up usage in US, Canadian, and British style guides (APA, CMOS, MWM. TCS, OSM, and NHR; the last two do mention that many British publishers space the en dash) is overwhelming. Like CMOS, I would usually prefer to or through for a complete date, but I certainly see no problem with Christmas Day–New Year's Eve when we can use New York–Chicago flight—there’s little ambiguity with either.
• Under (2), it might help to explain why Franco-British rivalry takes a hyphen rather than an en dash—perhaps something to the effect of “the order of Franco and British cannot be reversed”, or “Franco is a prefix than cannot stand alone”, or “Franco is not lexically independent”. The last would be my last choice, because although its meaning may be obvious to most of us here, it may not be to others, and searching for the meaning is no simple task.
• Under (3), I would add one example in which the first component contains a space (e.g., Chuck Berry–style lyrics, New Zealand–style clothing, golf ball–sized brain.
• The usage within items in a list (which I’m still not convinced is a distinct case) should not lend itself to be read as precluding alternatives, of which the unspaced em dash is the most obvious, but not the only possibility. If a style has been established for a specific class of articles (e.g., Days of the year), that style should be followed. For articles on discography, the format is not nearly so uniform, so any reasonable alternative should be acceptable. I don’t think we should give alternatives any more than a brief mention, and even then, it should only serve to indicate that there are alternatives. JeffConrad (talk) 07:15, 13 July 2011 (UTC)

Thanks Jeff. On each of your points in order:

• I understand about keeping a uniform order of presentation. The trouble is this. It's good to keep the order wrong–right, consistently in the examples (as I have done). But the explanation may be shorter and easier to follow in the reverse order. I don't see a difficulty in the case you exhibit above: the examples almost stand alone to make the point. Anyway, sure: such a refinement can be flagged for incorporation.
• On the spacing of range dashes, the task is to interpret the voting and associated discussion, not to return yet again to the style guides. We have moved on from that, and must stay moved on if we want to make genuine progress and reach a conclusion. On the specific point you raise, though, about Christmas Day—New Year's Eve being justified by analogy with New York—Chicago flight, note that the sense of the dash differs. In one it shows a range, in the other it shows a relationship between two elements. Given that difference in meaning, it is well that the dash should at least behave differently with respect to spacing. That is at least as arguable. In the end, I believe the position I have incorporated reflects the clear majority in the voting.
• For the prefix element "Franco", I agree entirely that things should be made explicit for it. I did that, then trimmed it out in pursuit of less wording. It can easily be put back, if people want that.
• You write: "I would add one example in which the first component contains a space (e.g., Chuck Berry—style lyrics, ...". But that was not put to a vote, so we have no evidence of consensus for it. These differ considerably from prefix cases, and they are treated differently by various style guides (for what that's worth).
• I have little to say about dashes in those lists. For me it's pretty arbitrary, and a much easier thing to sort out than the other principles. Let others have their say: but let it be efficient and conclusive dialogue.

Noetica^Tea? 09:22, 13 July 2011 (UTC)

• I tend to prefer “right/wrong”, but don’t have strong feelings, and would not even insist that the order be consistent overall if variations would make the individual points more clearly. But I would urge consistency between the text and any specific examples; either order would probably work, as long as the order of the example matches that of the text that describes it.
• On spacing of range dashes, it obviously depends on how one interprets the voting. Though there seems to be more support for spacing, I obviously do not interpret the voting as clearly precluding closed-up use, as is supported by the overwhelming preponderance of widely respected style guides. I once again urge correct use of the singular and plural: you clearly speak for yourself and a fair number of others, but you definitely do not speak for everyone here, as the use of we would seem to imply. I am quite honestly baffled by “We have moved on from that”; if this is to say that we care not about widely accepted practice, it is almost as if we have extracted a position from a region in which the “sunny 16 rule” is inapplicable, and the position accordingly is little more than WP:ILIKEIT. I agree that there may be a difference in the sense of the two examples, but the difference is slight, and in any event, the mere existence of a difference does not in itself justify differential treatment. The issue is rather, I think, one of whether the closed-up usage leads to ambiguity—and it does not. As I previously indicated in my comment on the voting, we have essentially an issue of operator precedence, in the order of hyphen, space, and en dash. If we accept this, there is no ambiguity; if we do not, much of the overall rationale for using en dashes is weakened. Lest I be thought a pusillanimous pussyfooter, I can assure you that I will oppose a draft that proscribes the closed-up use.
• As for golf ball–sized brain, I agree that we did not vote on the specific example, but we did previously vote on the more general case of compounds that contain hyphens or spaces, and, by implication, that golf ball–sized brain would be preferable to golf ball-sized brain, even if the endorsement of the former was tepid at best. JeffConrad (talk) 10:23, 13 July 2011 (UTC)

Jeff, on the order of "right" and "wrong" examples and all that: no big deal. I suggest we defer adjusting such things, till weightier matters are dispatched.

On your remarks concerning the spacing of range dashes: For the third time, please desist from critique of my use of "we", which I have explained elsewhere. Stay with content, not its expression (on whose analysis we disagree). It seems that you refer to this statement of mine: "... the task is to interpret the voting and associated discussion, not to return yet again to the style guides. We have moved on from that, and must stay moved on if we want to make genuine progress and reach a conclusion." If you think our [sic] present task is different, please raise that matter in the subsection below: #Discussion of procedure. Meanwhile, a close quantitative and qualitative analysis of the voting and comments on the point at issue would be relevant and valuable. I hope you will undertake that, and report it here. Finally on this, you say: "I can assure you that I will oppose a draft that proscribes the closed-up use." Please note that if we all take such stands even before such shared examination of voting, and also elevate clearly minority positions to sub-guideline status, we will end up with unmanageable complexity. The guidelines will be unworkable: they will fail to guide. With that in mind, I have suppressed my own strong disagreement with ex–prime minister, though dissent from that in the voting would support my acting on that disagreement here. We must all be careful not to fall victim to WP:ILIKEIT. Everyone must compromise; that's how it is, with manuals of style.

On forms like golf ball–sized brain, if you thought the community should consider those, the time to raise it was six weeks ago when the points for voting were settled. We have moved on! :) Noetica^Tea? 22:48, 13 July 2011 (UTC)

The points for voting were "settled" overnight without discussion from a list intended for a different purpose. They omitted much; this is one point that was omitted. As for we, Noetica does not speak for anyone but xemself; unless this is a case in which Mark Twain is correct. ;} Septentrionalis PMAnderson 23:04, 13 July 2011 (UTC)

I don’t necessarily object to we; I use it here myself, but not in the sense of purporting to speak for Wikipedia when that is not necessarily the case. You feel that we have “moved on” from widely recognized sources, but clearly not everyone agrees (though I would certainly not go so far as to say that you speak only for yourself). You’ve also offered a definition of “consensus” that differs markedly from WP:CONSENSUS. I don’t necessarily see a “clear majority”, and it’s never been been strictly a matter of numbers, anyway—we seem to be missing the “quality of arguments”. If practice supported by the preponderance of quality guides is to be blown off, especially without any real reason for doing so, what then constitutes “quality of arguments”?

If you’ve followed my comments, I’ve generally been very supportive of what’s been proposed, but I now find myself almost to the point of clearly recognizing that the MOS is a guideline, and using my own judgment for recommendations are little more than dogged insistence on WP:ILIKEIT, especially when they represent minority practice. I assure you that I shall not be writing golf ball-sized brain (though I concede the opportunity to do so may never arise). JeffConrad (talk) 02:35, 14 July 2011 (UTC)

My comments:

• I like Kotniski's draft more. (This one is a bit too wordy.) :-)
• I'd always use {{xt}} and {{!xt}} in ways that it is obvious from the surrounding texts which are the right examples and which the wrong ones, even without relying on the colour or on symbols such as asterisks, as in “right, not wrong”, or “avoid wrong; prefer right”. But then, I once proposed for {xt} to have a tick and underlining and {!xt} to have a cross and strike-through, but it was opposed as too much clutter. A darker red and a paler, more bluish green (e.g. maroon and teal) would be easier to tell apart for colour-blind people, but for readers with non-standard display gammas the former might be too dark (harder to tell from regular black text) or the latter too pale (harder to read against the background).
• IMO, image filenames are even less relevant than recasting or hard spaces or exception would be (for these reasons).
• As I already said elsewhere, I'd suggest the advisability of recasting the phrase using different examples than the ones where a dash is suggested. I'd use examples like the current ones for the dash (with such über-familiar noun phrases as prime minister or World War II), and then after “... is possible and better” show seriously abstruse examples in red followed by a suggested rephrasing in green.

― A. di M.​_plé​^dréachtaí 11:38, 13 July 2011 (UTC) ― A. di M.​_plé​^dréachtaí 11:38, 13 July 2011 (UTC)

Examples of what you have in mind for the last point would be helpful. Septentrionalis PMAnderson 21:55, 13 July 2011 (UTC)

A di M:

• I like Kotniski's draft too; and Kotniski likes the present draft (too). On wordiness, please specify which words you would remove from the present draft. In introducing it, I predicted that editors would make such suggestions. But again: we have seen how brevity is misread in this most complex of punctuation topics; and we must preempt abuse of MOS guidelines that would capitalise on lacunae.
• On the practical red–green issue, it may be best to deal with that another time. I like your comments. Meanwhile, we have a simple solution with "*" to get us through the present business.
• On image filenames and the like, thank you for attending to this matter of general provisions versus local provisions. I'm not convinced yet on this one; I had thought that hyphens and dashes in filenames was a pretty confined issue. I wonder what others think?
• On recasting, I strongly suggest that we stick with rewordings of the actual examples given; but I also strongly favour using less familiar examples to start with. I did not include those for this point, because I wanted editors to recognise the examples – for continuity with earlier versions of the guideline, for this discussion. Let's alter them later, sure.

Noetica^Tea? 23:15, 13 July 2011 (UTC)

Answering the points in order: 1. I'm not sure exactly what words are the culprit, but I'm under the impression that Kotniski's version manages to say pretty much the same things more concisely and no less clearly (though a couple extra examples might be added); 2. I still think that not would be a lot clearer than an asterisk, though it works better if right examples are before wrong ones (“foo, not bar” vs “not bar, but foo”); 3. I think whatever argument for or against the avoidance of dashes in file names would apply to pretty much all other hard-to-type characters; anyway, I think it's pointless as most files are hosted on Commons so the uploader might have never even heard of en.wiki's MOS; 4. if we say that rewording is sometimes (but not always) better, I think we ought to include both examples of when it's better than when it isn't. As for the former, I recall someone asking on this talk page how to punctuate some very complicated technical term itself including a several-word technical term as a modifier, and someone else (IIRC either you or Tony) suggested a recasting. ― A. di M.​_plé​^dréachtaí 14:59, 14 July 2011 (UTC)

Thanks, Noetica, it looks like a good step, very clear and logical, lots of on-point examples. A few comments:

I am a little bothered by the statement "These can be quite similar to uses of the hyphen; or less often, the slash." The "quite similar" doesn't quite capture the right relationship. As one commenter pointed out in the voting or discussion, "interpretation is needed" in deciding whether some of the described roles are applicable in a particular case. Your examples make it clear that to decide which punctuation is needed in something like "blue–green", one has to consider the role, which requires interpretation; the process is not mechanical, because it signals the writer's intent; and that's why it's useful. Anyway, "quite similar" seems to sweep that point away. It would seem important to point out that the en dash roles differ from the roles of a hyphen; maybe something like "These are analogous to uses of the hyphen, which binds more strongly than the dash; or the slash, which separates alternatives."

Another point: I still think the item "4. To separate items in certain lists" is not a separate role, but just another case of using spaced en dash as a dash; same idea in Kotniski's point.

I'm surprised the "ex–prime minister" item causes trouble, but on review I see that several people had various problems with it. I note that it's taken directly from this guide by a Canadian, along with the "anti-xxx–pro-xxx" example, which would certainly not be OK with a hyphen, but for other reasons as you note; it's similarly expressed in lots of other guides, though often as an option for improving clarity "when needed." And this Adobe (American) book suggests "pre–World War II", and states that the hyphen is "incorrect" in that context; and on "post–World War II" and "San Francisco–based," we have the word of Chicago; many more books say similarly. So is this one more optional than most? If some editor changes "ex–prime minister" back to the hyphen version, should we complain and fix it back? I still think we should, since if an editor finds it useful to use punctuation to signal the structure more clearly, we ought to let them. In that sense, I don't see the point of trying to make this one more optional than most, as some are suggesting. On the other hand, I don't have strong feelings about it, and would be willing to go with adjustments if there's a consensus to do so.

I would support the idea of removing mention of filenames, since it is not a style issue, just a source issue.

On the spaced vs. closed-up en dashes, I'm happy to see that there's less of the spacing suggestion than before, and I agree that it should stay in the case of the date phrases with spaces in them, as it appears in every WP bio article. It's a step in a good direction (away from "too British for some"), and stops short of going too far; it's exactly as I would have interpreted the discussion.

Generally, I think you addressed most of the disagreements – even the "facepalm" one. It would be hard to address the issues of the two guys who disagreed with most of the en dash roles, and still call it a consensus result. A few more may show up now, like the one who complains that he can't tell a hyphen from an en dash in a typewriter font, so it doesn't matter what you use; but I agree it's time to try to adjust the dash section as constructively as we can given all the constructive feedback that has come out. Thanks again for this draft, building on Kotniski's. Dicklyon (talk) 03:54, 14 July 2011 (UTC)

North American and British practice differ on spacing, and on use in spaced compounds—NHR states that there is no satisfactory way of dealing with “ex-Prime Minister”, and notes (but does endorse or reject) the US practice (which would be “ex–Prime Minister”). The recasting seems preferable for a nominal (as shown), and I don’t really have a problem with “former Prime Minister Gordon Brown”, though I might just as likely write “former–Prime Minister Gordon Brown”. Though I have a bit of an issue with “San Francisco-based company”, the caps do help (if not as much as in “ex-Prime Minister Brown”). I probably can also sort out “pre-World War II technologies”; here, I think the major issue is how many words an open compound can contain before it becomes unmanageable (and where, if at all, might we want a nonbreaking space. I think it’s a bigger issue without the caps, as in “pseudo-page transition” or “golf ball-sized brain”. With the latter, I can sort out what’s meant, but not without a double take; because the former is nor written in English anyway, it takes several double takes (but perhaps this would be true for any concept with which a reader is unfamiliar). In short, I think the dash vs. hyphen a fairly big deal when there are no caps.

TCS does not seem to address the issue of open compounds, though they suggest an entirely different approach in some cases (e.g., red colour-filter [at 2.02(b)], apparently with the assumption that colour filter is a noun phrase).

I agree that it′s sometimes a matter of eye rather than rote; for example, I would probably write high school student, even though I’m unaware of a closed-up form of “school student” such as there is with “schoolteacher”. It may also be worth noting that some guides (e.g., APA and CMOS) do not use hyphens or dashes with ethnic identifiers (e.g., African American Student rather than African-American student. But this is an issue for WP:HYPHEN rather than here; I assume that we would prefer African–American relations regardless of how the ethnic association is handled. JeffConrad (talk) 05:59, 14 July 2011 (UTC)

@Dicklyon: I've heard that using dashes in compounds which would normally use hyphens when one of the words has a space is uncommon in BrE. If that's the case, the fact that it “signal[s] the structure more clearly” isn't enough to deprecate hyphens in this situation throughout en.wiki. For example, spelling the method of payment as ‹cheque› is clearer because it signals that you don't mean any of the other meanings of check, but you don't want to recommend that in articles written in AmE. ― A. di M.​_plé​^dréachtaí 11:16, 15 July 2011 (UTC)

It finds support in OSM (at 5.10.9), though the support is tepid at best. OSM does not endorse the hyphen as an alternative. As is discussed under What consensus means, the Fowler brothers saw the problem with using hyphen as early as the beginning of the 19th century (and had no solution by 1931, either).

One approach would be to permit either usage, possibly encouraging consistent use of either North American or British practice, though this could be tricky because some practices may be chiefly (but not exclusively) North American, and others may be chiefly (but not exclusively) British.

I can’t say whether the fact that one construction may indicate the structure more clearly is sufficient to deprecate an alternative construction, but it seems to me that the point of en dashes in nearly all cases is to more clearly indicate what is meant than would a hyphen. So if it doesn’t matter here, it arguably doesn’t matter anywhere. JeffConrad (talk) 22:39, 15 July 2011 (UTC)

Though Noetica and I seem to differ fairly strongly on what has been concluded with regard to spacing of range dashes, it should be clear that this is the only point in the draft on which we really differ. Though I’m not sure we really differ on the task that remains, we do appear to differ a bit on how this is to be done. I begin by looking at WP:CONSENSUS:

In determining consensus, consider the quality of the arguments, the history of how they came about, the objections of those who disagree, and existing documentation in the project namespace. The quality of an argument is more important than whether it represents a minority or a majority view. The argument “I just don't like it”, and its counterpart “I just like it”, usually carry no weight whatsoever.

I don’t mean to dwell too much on the IDONTLIKEITs, but do maintain that given this framework, what RSs have to say is very much relevant in determining what may have been concluded. I agree that more people seem to prefer spacing; whether it’s a “clear majority” is more subjective, but there certainly is some significant opposition. Accordingly, I again turn to the quality issue, and suggest that given the great number of RSs that support closed-up usage, it does not seem reasonable to proscribe that practice. I do not recall a single source being cited as the basis for requiring the spaced usage, though Butcher may have been mentioned earlier in the discussion. I think it helpful to review what she has to say on the subject:

En rules meaning ‘to’ and ‘and’ are usually unspaced: theocratic–military, chapters 11–9, 101–10. However, spaced en rules may be used between groups of numbers and words to avoid implying a closer relationship between the words or numbers next to the en rule than between each of these and the rest of its group:

6.6–8	but	6.6 – 7.8
September–January		18 September – 19 January
1215–1260		c. 1215 – c. 1260

But these spaced en rules should be used cautiously, especially if there are also parenthetical dashes, as the reader may not be able to tell one from the other; and it may be better to substitute ‘to’ in such cases.

Under “Parenthetical dashes”, she continues, “Spaced en rules are now most often used”, so it apparently is assumed that the parenthetical dash is indistinguishable from the spaced range dash.

Though I probably have more British style guides than the average Brit, I’m certainly not as familiar with them as I amd with the American guides. But of all the guides with which I am familiar, Butcher seems one of the most favorable to spacing range dashes. And the recommendation is far less than what’s current or proposed here. In particular, I simply do not see why a problem would arise with “Christmas Day–New Year’s Eve” when there is not a problem with “New York–Chicago flight”. In both examples, the dashes in both examples seem to me to be in the sense of to. And isn’t the association addressed by using the en dash rather than a hyphen? If not, much of the rationale for using en dashes elsewhere would seem to break down.

A possible compromise would be to include wording similar to that above, perhaps adjusting the sense of the dashes to match whatever may be agreed to. Clearly it’s impossible to arrive at formulaic guidance on when association implied by the en dash might misleading, and the same is probably true for any possible confusion between range dashes and parenthetical dashes. Practically, the choice would be a matter of personal choice—British practice aside from OUP is apparently to prefer the spacing, while North American practice is usually to prefer closed-up usage. The case for spacing is probably strongest with full dates and with instances similar to the last example above. Even so, NA seem to prefer closed-up usage. The most common instance of full dates is arguably in obituaries, and unfortunately, they aren’t always much help. If my local newspapers are any guide, the copy editing is invariably done by the deceased. JeffConrad (talk) 09:38, 15 July 2011 (UTC)

Though my main issue here is with Christmas Day – New Year’s Eve (it’s really not different from New York–London flight, and calling a range of dates is stretching things to the limit), I overlooked another: 28 mm – 17 m. The en dash could initially be confused with a minus sign. There really is no ambiguity, just as there isn’t with closed-up dashes in most date forms (at least for North Americans), but the reader may need a double take to see what’s really meant. The implication of minus is far less with closed-up usage, but since there seems to be reasonable consensus against closed-up usage when both ends of a range of dates, times, or similar contain spaces, closed-up usage probably is not an option. It would be better to write 18 mm to 17 mm. This is consistent with CSE6:

When expressing a range of numbers, use the word “to” or “through” to connect the numbers. Alternatively, an en dash, which means “to”, may be used but only between 2 numbers that are not interrupted by words, mathematic operators, or symbols to avoid confusion with the minus symbol.

This is stated somewhat better in the Alaska Department of Fish and Game Writer’s Guide (which cites CSE). Perhaps Ms. Palin could in turn cite it in her debate with President Obama when mixing units in a range (“3 moose to 4 caribou”).

This usage really wasn’t put to a vote (and got little discussion there aside from A. di M.’s comment), which is perhaps why the potential confusion wasn’t noticed. Incidentally, I would replace Obama–Palin debate with something else, because it quite likely will not happen, and consequently might confuse some readers. I’d return to Lincoln–Douglas debates, the stature of which Nixon–Kennedy debates cannot come close to matching. There is no potential confusion with a style if the plural is used. JeffConrad (talk) 01:40, 16 July 2011 (UTC)

Dunno, a sane person would typically write that as −(17 m − 28 mm) if that's what they mean. Also, in most cases it will be completely obvious from the context that you're looking at a range of positive numbers (so long as you don't use en dashes and minus signs in the same phrase, as in the red examples currently in the MOS). ― A. di M.​_plé​^dréachtaí 12:13, 16 July 2011 (UTC)

A slightly different issue, I think. The spaced en dash in most cases is not obviously different from a minus sign, and the first-impression interpretation of a spaced minus is subtraction, in this case “28 mm minus 17 m”. The context will usually prevent ambiguity (I don’t think we normally add or subtract wavelengths), but comprehension may require slight additional processing—the first impression I have is that of an operation involving units of length, which is plausible. The claim of real ambiguity is in most cases nonsense; no one is likely to mis-associate the numbers and words in something like “24 June–17 July”, though again, perhaps a brief rescan may be needed. One could also maintain that “3 June 1970 – 9 September 1971” is potentially ambiguous because of implied subtraction “1970 − 9”, but at least for me, that first impression is less likely because I don’t normally think of subtracting dates. To me, the most jarring unspaced construction of dates would be something like “3 June 1970–9 September 1971”; the first thing I see is “1970–9″, which could mean “1970–1979”. Again, I quickly sort it out, but would prefer not to need to do so.

Incidentally, spacing en dash in dates in at least one construction introduces genuine ambiguity. Consider “18 July – 3 September 2012” does this mean “18 July 2012 through 3 September 2012” (the dates of the Obama–Palin debates) or “18 July 2011 through 3 September 2012” (the time between now and when these debates are behind us)? Without the spacing (“18 July–3 September 2012”) the close association of the months strongly suggests the former meaning. Though it’s hardly possible to definitively conclude that North American practice is “better” or “worse” than that elsewhere, I think it’s worth noting that with the former, there is little chance of confusion among the em dash, unspaced en dash, and (spaced) minus. JeffConrad (talk) 07:19, 18 July 2011 (UTC)

I definitely agree that the example Obama–Palin debate is unacceptable. And use the Lincoln–Douglas debates to clarify that you don't mean some generic Lincoln–Douglas debates. Dicklyon (talk) 04:51, 18 July 2011 (UTC)

What consensus means

Consensus does not mean "a clear majority"; if we ran by majority vote, we would say so. Please see WP:CONSENSUS for how we do run; we do not use the proper meaning of consensus as "unanimous consent", which frequently causes difficulties over international treaties, whem a handful of states decline to join an overwhelming majority. (For more on this, see Law of the Sea Treaty, and others.)

Even on Wikipedia, "[t]he goal of a consensus discussion is to reach an agreement about page content, one which may not satisfy anyone completely but which all editors involved recognize as a reasonable exposition of the topic."

This is not a reasonable exposition; a third of us object to requiring ex–prime minister, ranging from those like me that would permit it, to those who would recommend against it. There is also strong sentiment against the anti-conscription–pro-conscription debate; the weakly supported (but unopposed) recommendation of stronger language to avoid all such forms is omitted. Septentrionalis PMAnderson 16:45, 13 July 2011 (UTC)

Which is why I keep harping on the idea of phrasing things as "preferences". When there is disagreement among Wikipedians over a style issue ... (and a third objecting is definitely a "disagreement") we should not impose the majority view... instead we should present the options, indicate that all are "allowed"... however, we can then note that one style is preferred (to put this another way... a "preference" can be determined by a majority vote, a "rule" needs clear and solid consensus). Blueboar (talk) 17:02, 13 July 2011 (UTC)

Would you care to do a draft? Septentrionalis PMAnderson 17:43, 13 July 2011 (UTC)

Actually, no... I have never pretended to fully understand the specifics of the endash vs emdash vs a hyphen debates (hell, I am still not sure I completely understand what the difference between them is). Because of this, I am not really focused on what the MOS should say, but rather on how the MOS should say it.

I have been trying to find ways to break the "my way or the highway" deadlocks that those of you who do understand the nuances of dashes and hyphens keep getting into. Looking at the AfD, it is clear (even without fully understanding what is being discussed) that, on some points, almost everyone agreed. I think that consensus would justify presenting these items in terms of being accepted "rules"... conceptually telling readers: "this is how it should be done". However, it is also clear that on other points there was significant disagreement. I think that lack of consensus indicates presenting these points in terms of "preference"... conceptually telling the reader: "here are the options... all are allowed, but we prefer X". Blueboar (talk) 20:05, 13 July 2011 (UTC)

I'll see what I can do in 24 hours. Septentrionalis PMAnderson 20:24, 13 July 2011 (UTC)

The claimed majority is not merely not consesuus, but false. The discussion or ex–prime minister is at Wikipedia_talk:Manual_of_Style/dash_drafting#In_compounds_whose_elements_themselves_contain_hyphens_or_spaces; section 5b. At most 10 support requiring the dash, and of those Jeff, Tony, Kotniski, and MacWhiz are doubtful; 14 (including Ozob) would make it optional or deprecate it. Septentrionalis PMAnderson 23:53, 13 July 2011 (UTC)

To clarify my position on the use in compounds: as Fowler notes (TKE, under HYPHEN)

It is good English usage to place a noun or other non-adjectival part of speech before a noun, printing it as a separate word, and to regard it as serving the purpose of an adjective by virtue of its position . . .

He then proceeds to review the problems that arise with compound modifiers (mentioning his whimsical Anglo-SouthAmericans that he realizes is doomed before even being proposed). I think the en dash is a much better solution, but it can only correct so much of a mess—there is some practical limit to the number of these non-adjectival parts of speech that can reasonably be employed. But like consensus, I don’t think it’s strictly a numbers game. I would prefer recasting quasi-legislative–quasi-judicial to quasi-legislative, quasi-judicial, partly because of its length, but also because nothing is lost in recasting—in fact, I think quite the opposite obtains.

With the anti-conscription–pro-conscription debate, the choice is not so obvious, and as it stands may well be the least of evils. An option might be volunteerism—conscription debate, but this is somewhat less clear because the nouns are different (I had to think about the meaning for a moment). I think the main reason for my tepid support is that it’s less easy for me to quickly distinguish en dashes from hyphens than to distinguish en dashes from spaces. But perhaps this is just a personal quirk.

As for the prescriptivism–descriptivism debate, it should be clear that I lean toward the former (especially for en dashes), but recognize that when well-established alternatives exist (especially when they reflect ENGVAR or its equivalent), I might allow any of them, possibly preferring one, if that alternative would normally reduce ambiguity. Like Blueboar, I might prescribe (or at least strongly recommend an approach if it’s one on which we clearly agree (e.g., parenthetical dashes and en dashes in ranges of numbers), especially when that practice finds nearly universal acceptance outside of WP. One always must be careful about claims of universality, of course; some guides (SP 811) deprecate range dashes for numbers in technical contexts, and some organizations (ASTM) proscribe them. In some cases (unspaced em dashes vs. spaced en dashes), an expressed preference would clearly not find any traction, and probably should not. And finding a justification other than WP:ILIKEIT would be almost impossible. One thing I certainly would not do without a mighty good reason is proscribe a common practice (e.g., closed-up use of range dashes) that finds broad support in the guides—it is difficult for me to see doing so as other than utterly capricious. The more such provisions we include, the more likely are editors to ignore the MOS (assuming they don’t already). JeffConrad (talk) 09:13, 14 July 2011 (UTC)

Thank you for reminding me of TKE, although I would ascribe it to both Fowlers. The section is available on line, and it is more concise that MEU in saying that there are only three forms of composition (closed up; spaced; hyphenated) and that the only call for a fourth is Anglo-South-Americans, where there is are different levels of composition. They recommend, as we should, Warsaw and Brest-Litovsk railway line instead of playing with hierarchy. Septentrionalis PMAnderson 02:19, 15 July 2011 (UTC)

Indeed, that 1908 Fowler's "King's English" predates any guides that I am aware of that talk about en dash usage. And it recommends things like "Covert-Garden-Market salesman" that modern guides eschew; the insertion of hyphens into compounds used as adjectives is deprecated when the compounds are proper names (not "Golden-Gate Bridge", but just "Golden Gate Bridge"). Dicklyon (talk) 02:31, 15 July 2011 (UTC)

Yes, the Fowlers call a hyphen a regrettable necessity; more were necessary in 1908 than are now (and more seem to have been necessary in Oxford than in San Francisco, even then); a compounding dash is equally regrettable and much less necessary. Septentrionalis PMAnderson 02:44, 15 July 2011 (UTC)

I have the 3rd ed. (1931), which ponders Anglo-SouthAmericans—progress of sorts, I suppose. Though it was completed after F.W.’s death, much of the content obviously derives from his efforts. As many have correctly observed, the Fowlers were fighting some battles that have long since been resolved. But it remains a valuable work, even one makes it only through the five principles that being Chapter 1. Unfortunately, even in the 3rd ed., TKE is of no help on en dashes—I mentioned it simply to show that the Fowlers were well aware of the problem. That we′re still having difficulty finding consensus on some issues suggests that even with dashes, the problem remains vexing.

As regards Warsaw and Brest-Litovski railway line: for a reference to the railway, this works. But a Warsaw and Brest-Litovski trip works less well, and clearly, Warsaw-Brest-Litovski trip does not work at all. What the Fowlers did not propose was Warsaw–Brest-Litovski trip, which seems to me to do the job. I concede that this usage wasn’t as enthusiastically embraced as some of the others, and it’s not the prettiest construction, but what are the alternatives? I suppose we could resolve it the way a mathematician would resolve default operator by using parentheses, as I have done in a few examples—namely Warsaw-(Brest-Litovski) trip. I somehow don’t see this getting very far. JeffConrad (talk) 06:56, 15 July 2011 (UTC)

The proper rewriting of Warsaw–Brest-Litovsk trip is trip from Warsaw to Brest-Litovsk, which offers no problem.

As a minor detail, the press report the Fowlers are criticizing seems to have confused the name of the latter city with its adjective; if we use this example, we should amend it.

But this presents another reason not to use such compounds where they can be avoided; they are inherently vague. Septentrionalis PMAnderson 18:35, 15 July 2011 (UTC)

If one is willing to forgo the adjectival usage, then of course this rewrite works. And it may be just as good. But were it unquestionably the “proper” rewrite, the Fowlers would probably not have bothered trying to deal with the adjectival form.

I guess the vagueness is in the eye of the beholder—I see no ambiguity in the form with the en dash, though I agree with many who though recasting may often be better. But in some contexts, recasting may not be practical, so it’s nice to have a way of handling the adjectival form. JeffConrad (talk) 22:06, 15 July 2011 (UTC)

The dash in Warsaw–Brest-Litovsk can mean several things, depending on context; the Fowlers dealt with a case where it did mean and; they recommended (as the surviving Fowler continued to recommend) abandoning the effort to pack it into a single adjective. Where's the profit in it?

But the fact that the dash can mean almost anything means it is inherently vague. Sometimes the ambiguity is settled by context; sometimes it isn't. Septentrionalis PMAnderson 23:18, 15 July 2011 (UTC)

Here's something that's just occurred to me. For prescriptivism vs. descriptivism, if prescriptive points must be supported by reputable sources, such as style guides, then descriptive points must be supported by sources of similar reliability, such as sociolinguistic studies. (Come to think of it, many of the style guides cited as prescriptive have passages that describe actual use, so perhaps they'd be usable for both.) Personal observations of "what people do" should be dismissed as OR unless also so supported.

And Noetica is a dude. He says so on his talk page or at least he did when I made a two-second pronoun check three years ago. It's not the most relevant of issues, but can we stop asking about it now? Darkfrog24 (talk) 02:49, 15 July 2011 (UTC)

I would happily defer to "sociolinguistic studies" if that means a controlled experiment to determine whether typical readers, not just punctuation mavens, understand material more quickly when dash rule number 89-4C-6 (f) is followed. But only if such studies exist. Art LaPella (talk) 04:16, 15 July 2011 (UTC)

Darkfrog, I largely agree. I clearly lean toward style guides, more out of practicality than ideology. It’s usually a fairly deterministic process to find (and perhaps cite) what Guide A recommends. There is of course the problem that Guide B may recommend something different, but at least is also fairly straightforward to find the differences and decide to prefer one or just allow both if A and B are widely respected. We can always argue whether a given guide is “widely respected”, but most of the guides provide help by listing most of the other major guides in bibliographies. Now perhaps this is all just a vast conspiracy, but the guides listed by the other guides seem to correspond well to measures such as ASR (Amazon sales rank). Perhaps this is a sort of descriptive prescriptivism.

The alternative is far less straightforward. As Art indicated, a study relevant to a particular issue would need to exist, be discovered, discussed, possibly interpreted as to how well it relates to the issue at hand and as to what it actually concludes, as opposed to differing WP:ILIKEITs. This seems to me a tall order, and arriving at a conclusion could make the discussion on dashes seem brief and trivial by comparison.

That’s why I look to the guides—though this may not be a perfect solution, it seems simpler, faster, and less subjective than the alternative. JeffConrad (talk) 07:20, 15 July 2011 (UTC)

So do I; most guides do not recommend most of this, and the exceptions have low ASR. [signing comment by PMAnderson JeffConrad (talk) 22:13, 15 July 2011 (UTC)]

I guess we’d need to focus on specifics, but between APA and CMOS, most seem to be covered. And in the UK, NHR has reasonably high ASR. JeffConrad (talk) 22:13, 15 July 2011 (UTC)

Procedural subsection

¿There is still only halting progress because there’s complaining about what “consensus” means? The need to dwell on this point on Wikipedia is typically borne out of a minority side’s disputing that a consensus exists. Believe me—I know. I was exposed to a metric butt-load of Iranian-centrifuged, weapons-grade bullonium with complaints about how “There is no consenus!!!” over date de-linking and deprecating routine use of the IEC computer prefixes like “kibibytes”. In the end, the hold-outs lost because fighting against the consensus view is futile. What we need is the following:

1. All parties stipulate that Wikipedia:Consensus discloses all that is known and needs to be known about the subject.
2. Establish a clear consensus.
3. Find an uninvolved, respected admin who has huevos.
4. Have said admin rule on the consensus (win, lose, no consensus for the change).
5. Decisively deal with the holdout who has been giving you grief.

Greg L (talk) 22:39, 14 July 2011 (UTC)

The funny thing about consensus though is that it doesn't actually require anyone to "rule" on anything, "huevos" or not. Regardless... there is a well trod path through Wikipedia's dispute resolution process, what there is of it, which does utilize "finding an 'uninvolved' admin with 'huevos'" in order to adjudicate a ruling. I guess that it's just a pet peeve of mine that people refer to that (mistakenly) as "consensus", but... meh.
— V = IR (Talk • Contribs) 01:49, 15 July 2011 (UTC)

The sign of having achieved consensus is that you don't need an admin, nor huevos; nobody objects. That's what WP:CONSENSUS says; who objects to it?

Shouldn't this thread be under Procedure, the next section down? Septentrionalis PMAnderson 02:07, 15 July 2011 (UTC)

Note, I said thread, not section; I have made it a subsection. Septentrionalis PMAnderson 02:41, 15 July 2011 (UTC)

What consensus means

In the arbcom motion that led to this dash discussion, we were instructed "The discussion should be of sufficient structure to allow easy quantification of consensus rather than a large amount of poorly-framed debate." Casliber structured the discussion to make it easy to quantify agreements and disagreements with the various detailed provisions of the dash section. The result was that most sections had overwhelming support and agreement, but there were also a few less certain provisions and some constructive ideas for improvements. People who offered ideas and disagreements should check what they said, compare what ended up in Noetica's draft (the only currently active proposal, I think, since it has been largely accepted as an improvement on Kotniski's), and see if they're happy with it, and speak up one way or the other, so we can get the consensus finalized.

Jeff Conrad and I and a few others have weighed in above. I pretty much agree with Jeff that spaces around connective en dashes are rare (esp. in AmE), so I'm glad that the new draft backed off some from what the old one said. And I'm OK with the present compromise ("always unspaced, except for times and dates (or similar cases) when the components already include at least one space") as being motivated by the discussion (see 6b), even if it's hard to see exactly how far it should be changed from the votes at hand (PMAnderson is correct that the voting did not address this question, and the expressed disagreements leave a range of possible changes, of which Noetica's is somewhere in the middle); tweaking it can be left for a later round. At this point, I think minimal changes from the stable version, as motivated by a great discussion with wide participation, plus better clarifying examples, is what we can call consensus based on the discussion. Going further will require more work and more discussion to get to a consensus to do more, unless someone soon suggests specific changes that get little pushback. I made a few suggestions myself, but am content to let them slide to a later discussion, or forever. Let's get this done. Am I right? Or do we have to discuss a lot more first? Jeff? A. di M.? Kotniski? Tony? Anyone else who wants to comment on the new draft? Dicklyon (talk) 04:45, 18 July 2011 (UTC)

Discussion of procedure

[What to do next; when to do it]

I know probably no-one will like it, but here's my idea:

• From Day 1 to Day 7, whoever can be bothered to shall review all recent discussion about this topic, write a draft in their own userspace reflecting the consensus as well as they can, review and comment other people's drafts, and tweak their own draft addressing other people's comments.
• By the end of Day 7, I guess there will be a few such drafts (probably less than half a dozen); the drafts shall be locked down, and a new subpage shall be created where as many people as possible (through WT:MOS, WP:VP, WP:CENT, WP:RFC/STYLE, WT:WPMOS, maybe WT:GOCE, WT:TYP, WT:LANG, Talk:Dash, and possibly a watchlist-notice) shall be asked to review the drafts and sort them in order of preference, from Day 8 to Day 14.
• At the end of Day 14, the ‘ballot’ page shall be locked down and the preferences evaluated according to the Schulze method or similar. The ‘winning’ draft shall then be copied and pasted, verbatim, replacing the entire current Section 8.9 “Dashes” of WP:MOS.
• After that, tweaks to the guideline can be discussed and implemented through the normal WP:EDITREQ mechanism for protected pages.

(Of course, the better a drafter is at reflecting the community consensus, the more likely their draft is to win.)

― A. di M.​_plé​^dréachtaí 16:16, 13 July 2011 (UTC)

The Schulze method is contrary to our stated goals: to select the policy which as many people as possible can tolerate. Septentrionalis PMAnderson 16:28, 13 July 2011 (UTC)

I have been reviewing Noetica's draft, with respect to how well it incorporates fixes to all the disagreements that came up in the extensive discussions; I intend to recommend a few tweaks, but I think things would get out of control if I tried to make my own competing version; I don't think you'd find many editors with the patience to dig into differences and rank them at this point. But if someone thinks it is far enough off-base that they want to do a whole new one, I guess we'll look at that, too. As Noetica suggests, some notice of such an intention would help us keep our work in order. Sounds like there's some sentiment for a more "watered down" version from a few editors. Dicklyon (talk) 18:43, 13 July 2011 (UTC)

Only a few editors have commented on any draft.... Septentrionalis PMAnderson 18:54, 13 July 2011 (UTC)

Refactoring remains unacceptable; if Noetica repeats it, he or she (I don't know whicn) may explain that action elsewhere. Septentrionalis PMAnderson 23:53, 13 July 2011 (UTC)

Removed discussion (see preliminaries, above)

We now have an edit war over whether "What consensus means" should have === or ====. Would === = be an acceptable compromise? Another hot issue is Noetica's gender. See the top of his user talk page. And break your eggs on the big end only! Art LaPella (talk) 00:38, 14 July 2011 (UTC)

<chuckle> Fine; I was taught that gender should not matter; but Noetica was fiercely inovolved with the gneder-specific language movement. So rather than guess, I was non-specific.

What I object to, however, is having my comment refactorsd. That's obnoxious, and we have politcy against it. Septentrionalis PMAnderson 00:47, 14 July 2011 (UTC)

You presumably meant to say that only some comments should be refactored according to WP:TALKO (whether this is one of those times is an issue that you and Noetica can surely debate for weeks.) Art LaPella (talk) 01:17, 14 July 2011 (UTC)

No, I have never had a productive debate with Noetica (a neuter plural where I come from); I'd take him to ANI first. He does not have my consent. Septentrionalis PMAnderson 01:22, 14 July 2011 (UTC)

I did mis-guess Noetica's gender; this may (or may not) have to do with the fact that in my native language nouns such as matematica are feminine singulars. (But then, where I am from Andrea and Nicola are men's names, so I shouldn't use that as an excuse.) :-) ― A. di M.​_plé​^dréachtaí 14:45, 14 July 2011 (UTC)

WTF, nearly two days and neither of you guys has commented on whether it makes sense to call a girl Andrea given its etymology... :-) ― A. di M.​_plé​^dréachtaí 12:13, 16 July 2011 (UTC)

PMAnderson, you write: "Noetica was fiercely inovolved with the gneder-specific language movement [sic]." That is a coloured characterisation ("fiercely") of an involvement that I have no recollection of. Please retract it. Your irrelevant asides focusing on me are a distraction. You have failed to retract all your pointed innuendos. Please remedy that. And striking out text will not be enough; I would like to participate in the important discussion above without attacks and provocations, whether they linger in the text or are perpetuated in edit summaries. I'm confident that other editors would take this to be a perfectly reasonable request. Noetica^Tea? 03:07, 17 July 2011 (UTC)

I'm more inclined to assume good faith this time. I said "Another hot issue is Noetica's gender." That implied accusation entitles Septentrionalis to explain his half of that distraction. If, and only if, "Noetica was fiercely involved with the gender-specific language movement" is true, and if he has never been told Noetica's gender (although I've seen it repeatedly over the years by clicking Noetica's user name), then that is his explanation. Perhaps the most "inflammatory" comment above is my own: "... an issue that you and Noetica can surely debate for weeks", but somehow people seldom aim their best insults at me, on this page or any other battleground page. Art LaPella (talk) 06:13, 17 July 2011 (UTC)

Art, I admire your trust. But if you were as familiar with the relevant history as I am, that trust would melt away. PMAnderson knows exactly what he's doing. From one point of view, all of this is trivial. From another, it is of prime importance that we observe civility and work harmoniously. ArbCom has made the point explicitly. From the injunction under which we are currently working:

All discussions on the subject of En dashes in article titles discussion (interpreted broadly) are subject to civility and 1RR restrictions. Administrators are urged to be proactive in monitoring and assertive in keeping debate civil.

When PMAnderson acts and posts as he has, that is uncivil. When PMAnderson makes an edit summary that accuses me of "positive falsehood", that is uncivil; in fact, what he referred to in that summary represented merely a different take on certain voting, aided by a careless and slanted reading of what I had earlier said. You must know that I can look after myself in a street-fight, and have all sorts of tricks at my disposal if I choose to answer an assault with humour, incisive point-scoring, or ridicule. But the present matter is too serious for that. On the other hand, if people want to vary the framework and the suggestions I have made for the civil conduct of the discussion, sure: we could loosen up. Then we could have some real fun and fireworks. I'd rather we not do that; and I ask again that I be treated civilly (as the proposer of this move toward a resolution, with all of the difficult drafting that preceded it), as I treat others civilly also. I am keen to resume my involvement in the discussion; I'm waiting for conditions to be as ArbCom has said they should be.

Noetica^Tea? 06:47, 17 July 2011 (UTC)

Your sentence about "accuses me of 'positive falsehood'" is a legitimate complaint to the extent you explained it. But for those of us less able to comprehend the "relevant history" and all the "tricks at [your] disposal", you often seem to take offense at a long list of trivialities most of us wouldn't mention. In that case, emphasizing the need for civility counts against you. Art LaPella (talk) 15:45, 17 July 2011 (UTC)

So my insistence on the need for civility, which was also especially highlighted in an active ArbCom injunction, "counts against me"? I note your opinion, Art; others will agree with you, and others have agreed with me already. If I were to give the history that led us to this point, and has wasted weeks if not months of real time in people's real lives, you might find that uncivil of me. Hmmm. I say it would reveal incivility elsewhere; and it would reveal gross and sustained disruption to prove a point. I do not want to dwell on history, as I made clear in starting this whole section. Nor do I want to use kick-boxing techniques against unfair attacks. I can, but will not.

I am still waiting for a positive move toward respect and due process. Others may want to see the last several weeks' work dissipated; I do not.

Noetica^Tea? 22:21, 17 July 2011 (UTC)

Civility complaints are the last refuge of those with no case on the merits. Septentrionalis PMAnderson 23:15, 17 July 2011 (UTC)

I expected my carefully hedged suggestion that "you often seem to take offense at a long list of trivialities" might get an angry outburst, although that suggestion was a natural response to a demand for civility. But instead I got an objection to the hint that such an offense-taking tendency is in itself a kind of incivility, thus aggravated by demanding civility of others. Oh well. Finish the dash draft? Art LaPella (talk) 00:27, 18 July 2011 (UTC)

No need to hedge anything with me, Art. If you want to accuse me of incivility, do so directly so that we can settle things briskly and move on. You say: "instead I got an objection to the hint that such an offense-taking tendency is in itself a kind of incivility." But I was not hedging anything, and I did not mean or wish to hint anything beyond what I said: "I note your opinion, Art; others will agree with you, and others have agreed with me already." I want to avoid innuendos. Let's all do that. And let's be definite about all the issues here. Since my last post here, PMAnderson has come in with the gratuitous suggestion that in calling for ArbCom's injunction to be respected and enforced (for civility), I show that I have "no case on the merits". That would be laughable, if this were at all a laughing matter. I could respond in the same vein, and we would all get nowhere.

I have nothing to apologise for. Others do have something to backtrack on, and I am still waiting for the air to be cleared so we can continue our work without threats, arch attacks, or wanton misrepresentations.

Noetica^Tea? 00:52, 18 July 2011 (UTC)

I see no hope of re-explaining the re-explanation, so I'm giving up. Art LaPella (talk) 01:17, 18 July 2011 (UTC)

Art: If you take a little time away and then give the content of this section a detached reading, you may come to understand why I too have a feeling of the futility of it all. Then again, you may not. It's your call.

For my part, I am not giving up. Still just waiting. Noetica^Tea? 01:56, 18 July 2011 (UTC)

Summary; final request to complete discussion; and notice of a final draft

ArbCom's deadline (16 July) has passed. The only task now is to finalise a draft for the dash guidelines. Since it is imperative to conclude things in a collegial way and to bring this protracted matter to an end, I am setting aside my concerns over civility. I simply call once more for good will, restraint, and focus on the task as we wrap things up.

I propose to post a final draft, 24 hours from now. Someone has to, and since I have become intimately familiar with the task and its details, I propose to do this myself. I continue, of course, to suppress my own personal preferences in style; I hope that others will do the same.

Below is a summary of editors' points about the present draft, with provision for brief discussion. I really hope that editors will not prolong discussion unreasonably. There will be opportunities in future, to revisit any guideline. Simple, short statements like "Agree", or "I can live with that for now", would be extremely helpful at this stage. This is not the place for deep analysis, generalities, politics, or digressions. Please keep comments neat and suitably indented, and do not start or continue lengthy exchanges. I will consider everything in this new final stage of discussion, as I prepare a final draft. Again, someone has to. ☺

Noetica^Tea? 03:06, 19 July 2011 (UTC)

1. Length

Tony and A di M have taken up the matter of length that I raised. (I had written: "Editors will suggest ways to shorten it; but the voting has shown how easily brevity leads to entirely wrong interpretations. We need to guard also against intentional misconstrual.)" No one has suggested what might be taken out. A di M wrote: "I'm not sure exactly what words are the culprit, but I'm under the impression that Kotniski's version manages to say pretty much the same things more concisely and no less clearly." (I set aside the question of Kotniski's draft; I am among those who have thanked him for that effort; but the newer draft takes things further than his did.) It seems to me that if this were so, editors could find ways to trim the present draft. Let's have short, definite suggestions now, if anything is redundant and unhelpful in the draft.

Brief discussion

• Agree Fine, brevity. No brainer. Greg L (talk) 05:37, 19 July 2011 (UTC)
• Agree. Having read through it again, I'm impressed with the logic and efficiency of the examples. I learned from them, and other editors who want to should find them helpful and instructive. Editors who don't want to read them? Fine, others will be happy to fix. Tony (talk) 12:34, 19 July 2011 (UTC)
• Agree – The length is OK; lots of examples good; more examples not really needed; no problem if you make minor adjustments one way or the other. Dicklyon (talk) 14:58, 19 July 2011 (UTC)
• Disagree The length is excessive. Kotniski's draft is much shorter, and could be shortened further - yet it covers more ground. Septentrionalis PMAnderson 18:07, 19 July 2011 (UTC)
• Agree if that means the current length is OK. Brevity that results in ambiguity is of little benefit. As I discuss under Spacing in ranges, I erred in providing far too few examples, including one that eluded me until this afternoon when I actually had to use it. The proposal concentrates more on examples than abstract descriptions, which is as it should be. JeffConrad (talk) 04:29, 20 July 2011 (UTC)

Response in the final draft

• No specific actions were called for. I added a couple more examples, to satisfy other requests below. Noetica^Tea? 10:42, 20 July 2011 (UTC)

2. Content and layout of examples

A. Jeff Conrad commented: "The order of the recommended/not recommended practice in the examples should match that in the text that precedes it." I argued that this was not essential, thinking that the order among examples (wrong version, then right version) trumped other considerations. It's hard to meet both requirements and maintain clarity.

Brief discussion

• Agree Seems to be sensibly laid out. Greg L (talk) 05:37, 19 July 2011 (UTC)
• Agree, although I see Jeff's point. Tony (talk) 12:34, 19 July 2011 (UTC)
• Agree – the order is OK, and it would be hard to make it as consistent as Jeff suggests. Dicklyon (talk) 14:58, 19 July 2011 (UTC)
• A matter of rhetorical choices at each juncture; if the examples are clear without being made into a French garden, who cares? Septentrionalis PMAnderson 18:07, 19 July 2011 (UTC)
• Comment Recall that my suggestion was only that specific examples match the order of the text that describes them; It’s not a killer issue, but it would make it a bit easier on the reader. I see no need for overall consistence.
• Comment Unless there is good reason to do otherwise, I would like to see the examples show adjectival usage, because these are the more challenging usages. I give the example of the 20 kHz–20Hz  propagation delay, which fooled me until I just had to use it.
• Comment I would like to see all examples of ranges of dates (and perhaps of physical quantities with appended units) include nonbreaking spaces where appropriate, as called for in WP:Manual of Style (dates and numbers)#Non-breaking spaces (e.g., between date numbers and month names, as in 1790 – 1 December); new editors often learn by looking at the code. JeffConrad (talk) 04:51, 20 July 2011 (UTC)

Response in the final draft

• I have managed to satisfy Jeff's request regarding the order of texts and examples.
• Several examples already show adjectival use; I have added two more, if we include ex–prime minister Thatcher (Jeff accepts that as adjectival below, but I don't). It is important to show non-adjectival uses also; they can be trickier in their way. This suggestion would have been better made during the last week, while the now superseded draft was on the table.
• It was discussed earlier that the issue of hard spaces would be removed, for enhanced general treatment elsewhere in MOS. Again, it is awkward to have to revisit this issue at the last minute (beyond my 24-hour period of notice, and four days after ArbCom's deadline). For every reader of MOS who is helped by seeing &nbsp; in examples, there will be another is flummoxed and alienated by it. Best treated specially elsewhere; and best to revisit this issue another time. Noetica^Tea? 10:42, 20 July 2011 (UTC)

B. A di M commented: "I'd use examples like the current ones for the dash (with such über-familiar noun phrases as prime minister or World War II), and then after '... is possible and better' show seriously abstruse examples in red followed by a suggested rephrasing in green," and later: "... if we say that rewording is sometimes (but not always) better, I think we ought to include both examples of when it's better than when it isn't." My own view is that the present examples work well to make their points. I put a lot of thought into their selection. And we don't want more examples, surely.

Brief discussion

• Agree They look like good examples. Greg L (talk) 05:37, 19 July 2011 (UTC)
• Agree—No longer if possible, please. Very well laid out, IMO; it contains some tips for updating the rest of MoS in the future along these lines. Tony (talk) 12:34, 19 July 2011 (UTC)
• Agree – a few more or less examples won't bother me, but I see no problem big enough to get hung up on. Dicklyon (talk) 14:58, 19 July 2011 (UTC)
• Immaterial; but I'd like to see what A di M has in mind before deciding. Septentrionalis PMAnderson 18:07, 19 July 2011 (UTC)
• Agree with current examples, but might consider a few specifics if they would illustrate points that would otherwise be lost. JeffConrad (talk) 04:51, 20 July 2011 (UTC)

Response in the final draft

• I believe I have understood A di M's suggestion; I have added an example in response. Noetica^Tea? 10:42, 20 July 2011 (UTC)

3. Expository points

A. Jeff Conrad thought "it might help to explain why 'Franco-British rivalry' takes a hyphen rather than an en dash." Agreed. I can do that, and had simply omitted the explanation to get fewer words. Any similar points, anyone?

Brief discussion

• Agree Jeff gets his suggestion adopted. Greg L (talk) 05:37, 19 July 2011 (UTC)
• Agree Tony (talk) 12:34, 19 July 2011 (UTC)
• Agree – I'm OK with a bit more explanation, but don't see it as necessary really. Dicklyon (talk) 14:58, 19 July 2011 (UTC)
• Disagree. The explanation is that compounds normally take hyphens (or spaces, or nothing). We do not need to deal with exceptions to exceptions. Septentrionalis PMAnderson 18:07, 19 July 2011 (UTC)

Response in the final draft

• Despite PMAnderson's dissent, the request appears reasonable. I have satisfied it. Noetica^Tea? 10:42, 20 July 2011 (UTC)

B. Dicklyon was "a little bothered by the statement 'These can be quite similar to uses of the hyphen; or less often, the slash.' " (See details above.) I accept this, and will make a change that expresses things better.

Brief discussion

• Agree Dick made a good point. Greg L (talk) 05:37, 19 July 2011 (UTC)
• Agree Tony (talk) 12:34, 19 July 2011 (UTC)
• Agree – Thanks for hearing me; I'll trust that you'll improve it a bit. Dicklyon (talk) 14:58, 19 July 2011 (UTC)
• Let's see what happens. Septentrionalis PMAnderson 18:07, 19 July 2011 (UTC)

Response in the final draft

• Done. Explained. More needs to be said about these combining forms (Franco, Sino); but to get things concluded now, I'll hold back my concerns till the next time we visit the dash section. Noetica^Tea? 10:42, 20 July 2011 (UTC)

4. Spacing in ranges

Both Jeff Conrad and Dicklyon had things to say about this. (So did A di M, but it seems to me that his concern over confusion with minus signs was adequately addressed. I am ready to be corrected, of course.)

Jeff wrote, for example: "Spacing of range dashes where one or more elements contains a space: either spaced or unspaced use should be allowed. The polling seems to indicate considerable support for allowing both." And later: "Though Noetica and I seem to differ fairly strongly on what has been concluded with regard to spacing of range dashes, it should be clear that this is the only point in the draft on which we really differ." And again: "Clearly it's impossible to arrive at formulaic guidance on when association implied by the en dash might misleading, and the same is probably true for any possible confusion between range dashes and parenthetical dashes. Practically, the choice would be a matter of personal choice—British practice aside from OUP is apparently to prefer the spacing, while North American practice is usually to prefer closed-up usage. The case for spacing is probably strongest with full dates and with instances similar to the last example above. Even so, NA seem to prefer closed-up usage."

Dick wrote: "I pretty much agree with Jeff that spaces around connective en dashes are rare (esp. in AmE), so I'm glad that the new draft backed off some from what the old one said. And I'm OK with the present compromise ("always unspaced, except for times and dates (or similar cases) when the components already include at least one space") as being motivated by the discussion (see 6b), even if it's hard to see exactly how far it should be changed from the votes at hand (PMAnderson is correct that the voting did not address this question, and the expressed disagreements leave a range of possible changes, of which Noetica's is somewhere in the middle); tweaking it can be left for a later round."

My own view, after careful consideration of this vexed issue: Jeff's concern is perfectly understandable. I had incorporated a "marginal" case of a compound date range (Christmas Day – New Year's Eve); strictly, it is such a case, and it would be a denial of the responsibility for a manual of style simply to exclude it as "too hard". People look to a manual to decide difficult cases. We could argue forever over this one, and about the principle as a whole; but in the end, we would still have to settle on a compromise. Dick is "OK with the present compromise", and "Noetica's [proposed change is] somewhere in the middle". Finally, I agree with Dick that "tweaking it can be left for a later round"; and that applies to a number of finer points. Let's settle things for now, not for Doomsday or the afterlife.

Brief discussion

• Agree A sensible amalgam of all input. Greg L (talk) 05:37, 19 July 2011 (UTC)
• Agree. This is one of these cases where no "solution" can satisfy every fiddly case. Noetica's compromise is appropriate. Tony (talk) 12:34, 19 July 2011 (UTC)
• Agree – It's OK, and I won't object if it changes a bit, now or later. Dicklyon (talk) 14:58, 19 July 2011 (UTC)
• Disagree The way to settle it for now, not for Doomsday, is to be silent. The only thing that can be said definitively out of this is that there are several ways to space here; we don't need to say even that unless there is some reason to "extend our legislative hand" (and so breach policy). Septentrionalis PMAnderson 18:07, 19 July 2011 (UTC)
• Partially Disagree. Though I, like the vast majority of guides, clearly prefer closed-up usage in all cases, my vote was a compromise, especially where we have a long-established practice of spacing the dash in complete date ranges in biographies. I still take issue with a few of the examples, however. In some cases, it is impossible to assess consensus because I don’t think we really voted on the issues; in a couple of others, spacing the dash creates real ambiguity rather than “initial ambiguity”, so it becomes far more than a matter of personal preference. Some specifics: I fail to see how different treatment of New York–London flight and Christmas Day–New Year’s Eve can be justified; there’s certainly no difference in “initial ambiguity”, and the latter is simply not a range of dates. The genuine ambiguity is more serious, as arises from 18 July – 3 September 2012. Even worse is ambiguity that arises spacing the dash in a range of quantities with repeated units. I just encountered a real-world example for which even my suggestion to use to would not work: I had need to refer to “the 20 kHz–20 Hz propagation delay”. Had I used the 20 kHz – 20 Hz propagation delay, I’d had have left the strong impression of subtraction, which would have been dead wrong because the actual difference here was one of time, not frequency. Now perhaps the reader would have sorted it out, but doing so would have required unnecessary effort. And I don’t think my suggestion of using to (the 20 kHz-to-20 Hz propagation delay) really works here, either. This example raises another point: unless there is good reason to do otherwise, I think the examples should show adjectival use, because it is usually the most challenging case.

Perhaps it could be said that these issues should have been raised earlier; if so, as the person who did the breakout, I must take much of the responsibility for failing to better separate some of the usages. Though some of my differences here are a matter of personal preference, I say once again that the sources agree with me. To the anticipated rejoinder that “we have moved beyond the guides”, I would say that I have a bit of a problem with summarily ignoring the “quality of the arguments” element of WP:CONSENSUS, especially since I raised it in polling. If we really feel that “I like it” sometimes carries the day, we should revise WP:CONSENSUS accordingly. If we simply ignore it, I fail to see how we could chide anyone for ignoring this guideline as well. Again, please take my comments here in perspective—recall that my first words were “It looks pretty good”, and the cases here are relatively few. Nonetheless, my disagreement remains strong on a few of them. JeffConrad (talk) 04:22, 20 July 2011 (UTC)

• OK. Well, why not try working offline (backchannel, on your talk pages, whatever) with Noetica to hammer out a better solution. None of this stuff is locked in stone. Consensus can change. This is a complex package of changes and consolidations and there are bound to be points that can be modified or added. In that spirit, I hope one or both of you come back with a tweak you can accept in the spirit of compromise so you, Jeff, can be a member of an across-the-board consensus. Greg L (talk) 05:04, 20 July 2011 (UTC)

Response in the final draft

• Dick, and Tony want to move on; PMAnderson wants silence; Jeff wants to expatiate at the eleventh hour; and Greg wants Jeff and me to get a room, where he can talk the legs off a chair and I can talk them back on again (count them at the end: there'll be five by then). Let's see what I can do ...
• Jeff, your vote was a compromise? Fine. So was mine (I'd want certain headings to have been in this fold also, agreeing with at least five major style guides and de facto practice at OUP). The majority appears to be happiest with spaced en dashes for ranges when either component has a space; so we meet in the middle, with them. Now, you say that Christmas Day – New Year’s Eve is not a range of dates? I say it is; and that its similarity to New York–London flight is as superficial as the similarity of blue~green algae and red~green colorblindness (which we distinguish with "-" and "–"). We need to identify Christmas Day – New Year’s Eve as a range of dates in the examples, to make a robust, simple guideline that anyone can understand. It does no more harm to space this one than it does to space any other date example (with spaces in it). For that reason I have added a further example: Christmas 2001 – Easter 2002. That is a date range, surely. The user gets the idea, and can remember and apply it. Bear in mind that these date ranges typically occur in parentheses, where they can do no harm to the enclosing sentence. As for your highly unusual and recently discovered example Jeff (the 20 kHz – 20 Hz propagation delay), remember that hard cases make bad laws. There will always be rare, fluky situations that call for creative solutions; we soon get entangled if we make ourselves hostage to them.
• Finally Jeff, you have misunderstand me if you think that I interpret consensus as only a matter of counting votes. I wish you had not raised this right here; but since you did, I will answer. I wrote of the "practical necessity" under which I operated, given Casliber's requirement they we vote clearly, to "quantify" the consensus. Take him to task, not me. It was extraordinarily hard to work up a draft drawing on every last eddy of opinion, in those two subpages crammed with comment. Luckily, the voting corresponded pretty well with the weight of argument and evidence (sign of a thoughtful group). I suggest that you revisit the top few sentences of WP:CONSENSUS, yourself. Take PMAnderson along with you. I quote: "Consensus is not necessarily unanimity. Ideally, it arrives with an absence of objections, but if this proves impossible, a majority decision must be taken. More than a simple majority is generally required for major changes." We have every reason to believe that this is such a case. Practical necessity, once more. But still, who is not trying for something better, by reason, compromise, and consultation? I know I have been, for the last several months. Noetica^Tea? 10:42, 20 July 2011 (UTC)

You are proposing a major change, with (even by your own count), a hairline majority of 15-14. By your own argument, this should not be done; silence would be preferable. Septentrionalis PMAnderson 16:48, 20 July 2011 (UTC)

• Greg, with all due respect, after reviewing the results here half a dozen times, I have difficulty finding a clear majority, and nothing remotely approaching an “across the board consensus”; because of the way many of the votes were qualified, I have trouble even arriving at an unambiguous tally. Given your earlier comment “All parties stipulate that Wikipedia:Consensus discloses all that is known and needs to be known about the subject”, I find your comment here puzzling—more than a simple majority would seem needed, and I cannot see that “quality of arguments” was even considered. JeffConrad (talk) 21:37, 20 July 2011 (UTC)

Noetica, I’m not sure I’m taking anyone to task. I simply mentioned that I have great difficulty seeing a “clear majority”, and that I still cannot see how “quality of arguments” was considered. As best I can make of it, only two of us gave much more than “I like it” arguments here. What are the reasons for spacing? And what are the five major style guides to which you refer? Perhaps I missed something, but I do not recall recent mention of any of them. And for what it’s worth, the OUP books that I have close-up en dashes, though concede that I probably don’t have enough for a representative sample.

I’d hardly call the 20 kHz–20 Hz propagation delay a fluke, though I admit that adjectival use probably isn’t all that common. I would say this: while the reader can probably sort out either closed-up or spaced usage, the latter at first glance resembles the common situation of a difference of physical quantities, while the former presents no such ambiguity, because a minus sign in subtraction is always spaced (except perhaps for those who write unreadable code). So the closed-up usage here avoids ambiguity, while not presenting other problems. Hard cases make bad laws? Perhaps in Australia . . . in the US, we recognize that the safeguards of liberty have frequently been forged in controversies involving not very nice people.

As for “eleventh hour” I simply cannot agree—I’ve simply restated issues that I do not think were addressed. You obviously fee otherwise; sometimes reasonable minds can differ. JeffConrad (talk) 21:37, 20 July 2011 (UTC)

So have I. Despite Noetica's persistent personal attacks, I have been willing to compromise; I supported Tony's original proposal, which gave the permission half of those polled now support. It failed because many editors found it too restrictive (if I recall correctly, it was 19-18, and that was held not to be consensus); to claim something even more restrictive as consensus (on a narrower vote) is not accurate. Septentrionalis PMAnderson 20:19, 22 July 2011 (UTC)

5. En dash with spaced components

Jeff Conrad wrote: "I would add one example in which the first component contains a space (e.g., Chuck Berry–style lyrics ...". And further on this matter: "One approach would be to permit either usage, possibly encouraging consistent use of either North American or British practice, though this could be tricky because some practices may be chiefly (but not exclusively) North American, and others may be chiefly (but not exclusively) British. / I can’t say whether the fact that one construction may indicate the structure more clearly is sufficient to deprecate an alternative construction, but it seems to me that the point of en dashes in nearly all cases is to more clearly indicate what is meant than would a hyphen. So if it doesn’t matter here, it arguably doesn’t matter anywhere."

Dick wrote: "So is this one more optional than most? If some editor changes 'ex–prime minister' back to the hyphen version, should we complain and fix it back? I still think we should, since if an editor finds it useful to use punctuation to signal the structure more clearly, we ought to let them. In that sense, I don't see the point of trying to make this one more optional than most, as some are suggesting. On the other hand, I don't have strong feelings about it, and would be willing to go with adjustments if there's a consensus to do so."

Myself, I articulated my objections to this principle in the voting. I think it is quite unnecessary; but I put it in the draft because including it seemed more acceptable to more people than leaving it out. More importantly, I think, a substantial majority have expressed a wish for a simple principle, here and throughout the voting: either explicitly, or more often by their simple "agree". That is my view also: I don't like it, but stability and consistency on this particular principle is important in article titles especially, so let's go for the dominant way to lessen discord in future, and also possible searching problems. I also pointed out to Jeff Conrad that cases like Chuck Berry–style lyrics were not voted on. Again, the present round of discussion need not be final; but it ought to be finalised. If anyone feels strongly about this in the future, let it be raised when the present work is behind us.

Brief discussion

• Agree Good analysis on how to handle en-dashes. Greg L (talk) 05:37, 19 July 2011 (UTC)
• Agree—No strong feelings, but I'm more used to it now, and have even introduced it in a recent FAC lead, because it seemed more elegant than a triple hyphenated unit. While I think some readers may not be familiar with the usage, it's intuitive enough. Tony (talk) 12:34, 19 July 2011 (UTC)
• Agree – I like this Americanism, but I won't object later if we want to adjust it to be more optional. Dicklyon (talk) 14:58, 19 July 2011 (UTC)
• Disagree. There was a majority against requiring ex–prime minister; there was a majority to perfer it; neither was consensus. Any draft which requires it deserves dispute resolution. Septentrionalis PMAnderson 18:07, 19 July 2011 (UTC)
• Agree – This works for me.--Epeefleche (talk) 20:32, 19 July 2011 (UTC)
• Agree Like Dick, I think it adds clarity in some cases, and probably more so than with prefixes. And as I’ve said, if a hyphen is OK here, it’s OK in many other cases where we call for an en dash. Be assured, though, that I won’t be running around looking for uses of the hyphen here to correct. I would, however, have the examples show adjectival use (e.g., ex–prime minister Brown) because I think the benefit is more obvious}} (and how should “former prime minister Brown” be handled?). JeffConrad (talk) 05:02, 20 July 2011 (UTC)

Response in the final draft

• My analysis of the voting differs from PMAnderson's. Analysing the votes with their associated comments (as everyone accepts that we should), and seeing what the voters were willing or unwilling to accept, I count it this way. Of the 29 who voted on this principle (including PMAnderson, though he did not vote outside the navbox):
  • 7 rejected it outright (SarekOfVulcan, Powers, Hullaballoo Wolfowitz, SJ, Headbomb, Nageh, Telpardec)
  • 7 accepted it, but asked that it be optional (WFC, A di M, Powers, Enric Naval, Ozob, JN466, PMAnderson)
  • 15 accepted it, and did not ask that it be optional (JeffConrad, Kotniski, Armbrust, Dabomb87, Tony, Kwami, ErikHaugen, Dicklyon, CWenger, Jenks24, Courcelles, macwhiz, Finell, InverseHypercube, Noetica [see my follow-up on the discussion page, and my remarks earlier here])

That's a simple majority (15–14) accepting it without any option; and that's a clear majority accepting it in one way or another (22–7). Only 7 out of 29 favour giving a choice. Nothing in WP:CONSENSUS obliges us to accommodate every minority preference in a manual of style. That's not what manuals of style are for. Last, we don't know what those 7 would say to this question: "Do you want a manual of style to set a simple uniform guideline for everyone?" That number would have been 8, remember: but I gave up my opposition to the principle for the larger good: consistency of style on Wikipedia. In fact, I would have held out and maintained opposition; but Jeff was so adamant, and the principle is so widely accepted in America, that I yielded.

• I have amended the heading for the section; we voted on prefixes, and it is confined to prefixes.
• By an extreme act of contortion, I have amended the guideline in the final draft so that it applies in an acceptable, well-delimited set of cases, and recasting is to be clearly preferred in all other cases. I submit that this is an optimal solution. Noetica^Tea? 10:42, 20 July 2011 (UTC)
  • In short, there is no consensus for making such forms mandatory. Any draft which makes it mandatory is absolutely unacceptable. A draft which expresses preference, as Blueboar suggests, would be acceptable to me and to at least three-quarters of those polled; it is conceivable that those who strongly prefer ex-prime minister, or avoiding this construction altogether, would settle for being left alone. Septentrionalis PMAnderson 15:26, 20 July 2011 (UTC)

6. In lists

Jeff Conrad wrote: "The usage within items in a list (which I’m still not convinced is a distinct case) should not lend itself to be read as precluding alternatives." And Dick wrote: " 'To separate items in certain lists' is not a separate role, but just another case of using spaced en dash as a dash; same idea in Kotniski's point."

Do editors really care enough about this right now, though? The present draft does the best it can to preserve what the current stable guideline says, but just makes it clearer. I propose that we leave this topic for possible consideration later, also.

Brief discussion

• Agree That’s fine too. Going back to serial tweaking later to further flesh this out (like we used to do) will kill no one. Greg L (talk) 05:37, 19 July 2011 (UTC)
• Agree. I must say, spaced en dashes in lists look the best to me—far better than most alternatives—but the draft is fine. Tony (talk) 12:34, 19 July 2011 (UTC)
• Agree – It's kind of a misfit point, but not doing any real harm; defer adjustments to later. Dicklyon (talk) 14:58, 19 July 2011 (UTC)
• Disgree This has never been important; nobody has argued for this at all strongly; Jeff has expressed continued puzzlement as to what this means or why. The provision this has evolved from was a profitless choice between two perfectly reasonable ways to format lists. Take out the whole point, until somebody presents an actual difficulty. Septentrionalis PMAnderson 18:07, 19 July 2011 (UTC)
• Agree – This also works for me, for the moment, and agree that further changes could be considered at a later point.--Epeefleche (talk) 20:32, 19 July 2011 (UTC)
• Agree that because the proposal includes examples, it’s a big improvement, provided that we don’t see it as precluding other punctuation. I still have doubts about whether this is a distinct usage, but this is a minor issue that can be dealt with later if it’s felt necessary. JeffConrad (talk) 05:06, 20 July 2011 (UTC)

Response in the final draft

• Nothing to do; nothing done. Noetica^Tea? 10:42, 20 July 2011 (UTC)

7. Points to deal with elsewhere, outside the dashes guideline

A di M expressed this view: "Image filenames are even less relevant than recasting or hard spaces or exception would be."

I am happy to take out the mention of dashes and hyphens in filenames, subject to there being no objection here.

Brief discussion

• Agree To make progress and break out of the current morass, there is no harm not addressing file names now; that issue can always be addressed in the far, far off future in Wikipedia time (next week in real time) if needed. Greg L (talk) 05:37, 19 July 2011 (UTC)
• Agree Tony (talk) 12:34, 19 July 2011 (UTC)
• Agree – Image filenames are not part of "style", but I don't care if they're in or out at this point. Dicklyon (talk) 14:58, 19 July 2011 (UTC)

Response in the final draft

• Done. We must follow up with changes elsewhere in MOS, for this and the other points that were to be moved outside the dash guideline. Noetica^Tea? 10:42, 20 July 2011 (UTC)

8. Any points that have been missed in the above

[Please add only if they really are important; no new points please – they can be addressed another time.]

• None Noetica, you’ve truly done a great job trying to capture the essential elements of everyone’s input to arrive at a distillation around which a general consensus can form. It is a magnificent effort in an all-volunteer, collaborative writing environment. Your heavy lifting here has not gone unnoticed. Greg L (talk) 05:37, 19 July 2011 (UTC)

Brief discussion

• Great job, Noetica. Tony (talk) 12:34, 19 July 2011 (UTC)
• Thanks – for taking on the brunt of the work, and dealing with the inevitable blowback. I support you, and want to get this done, so will not offer any more new complications at this point. Dicklyon (talk) 14:58, 19 July 2011 (UTC)
• Noted. Noetica's draft does not even represent the views of a majority, much less consensus. Those who approve of it approve of ignoring discussion, usage, and sources. Septentrionalis PMAnderson 18:07, 19 July 2011 (UTC)

• Quoting you, PMA: Noetica's draft does not even represent the views of a majority, much less consensus. What defines a consensus is being determined now; here. Perhaps the consensus will be that it is a good step. Perhaps the consensus view will be that it smells like a Frenchman’s armpits; we’re not done yet. Whatever you think of Noetica’s effort at consensus-building in light of past water under the bridge is irrelevant now. No one said working in a collaborative writing environment is easy. You had an opportunity here to make compelling oratory to sway group opinion. The tenor of your posts makes me not so sure you exploited the opportunity to the fullest. Greg L (talk) 20:39, 19 July 2011 (UTC)

It does seem odd that PMA picks mainly on the minor "ex–prime minister" provision now, given that the section Wikipedia_talk:Manual_of_Style/dash_drafting#In_compounds_whose_elements_themselves_contain_hyphens_or_spaces was not among those that he disagreed with; he says it "deserves dispute resolution," as if that's not what we're approaching closure on. And unclear what he means by "not even...a majority" if one-third of the people expressed some reservations and disagreements (including Noetica, himself, who is now trying to find a consensus compromise that isn't even his own opinion). And now PMA objects to mentioning the role of en dashes as separators in list items, but he didn't say so when he had the chance at Wikipedia_talk:Manual_of_Style/dash_drafting#To_separate_items_in_a_list (actually, if you "open" the "superceded" bit there, you can see that he said he agreed). Now he says this "has never been important; nobody has argued for this at all strongly," which might be true, but holding up resolution for a point that he said he agrees on, and that he agrees is unimportant, seems like just a stalling tactic. If he wants to propose changing either of these later, I might even agree with him for a change, but that's not where we're at. As for his statement that we "approve of ignoring discussion, usage, and sources," I just don't get what makes him say such things; just trolling, I guess. Also note that on Noetica's draft that we're discussing final changes to, PMA offered neither any criticism nor any suggested changes, but just procedural complaints; so to object now to Noetica's attempts to address all the suggestions seems at least counter-productive, if not in plain bad faith. Dicklyon (talk) 02:07, 20 July 2011 (UTC)

It’s time to stop dumping on PMA. His style tends to be a hard-to-ignore one-man band. We all know that. In the past, too much has been made of his opposition and he has—in my opinion—had influence wildly in excess of head count. There’s been back & forth finger pointing and accusations that the other side has been operating… uhm… *cheatfully*. The Good Ship Major Malfunction has sailed. PMA is just one voice. He’s stated his opinion: fine, thank you very much; time to move on. If a consensus can’t be made here despite his opposition, then there are one or more shortcomings in Noetica’s proposal that need to be ironed out in order to get more editors singing from the same sheet of music. Noses back to the grind stone please. Greg L (talk) 05:16, 20 July 2011 (UTC)

• Though I still have a couple of very strong disagreements about a few points (and possibly the procedure at arriving at them), I stress again that the overall effort is a great job. JeffConrad (talk) 05:10, 20 July 2011 (UTC)

Response from Noetica

• Thanks! Now can we just settle this damn thing? It's been going for months. I don't know about others, but I'm exhausted. Just compromise, yeah? I myself have strong disagreements with the draft I have made. But it works. You will not find a more considered, more complete, more sensitive solution to this hardest of problems in modern punctuation. Not on the web; not anywhere. And we did it together. Noetica^Tea? 10:42, 20 July 2011 (UTC)
  • I think the fact that you can't find it anywhere else strongly suggests that large parts of this problem did not need to be solved by wikipedia. Wikipedia is supposed to follow its sources, not synthesize them in novel ways. I have to agree with PMA that it would be better to say less and prescribe less. Quale (talk) 06:03, 30 July 2011 (UTC)

Final draft 1: please endorse, or comment

I present what I am calling "Final draft 1" for the dashes section. It responds to suggestions made in the preceding subsection, where I have now enumerated the changes and responded to all concerns. Changes are marked in this new draft by underlining, and in one case by striking through. To see the new draft, click on "Show", at the right:

Noetica^Tea? 10:42, 20 July 2011 (UTC)

Suggested procedure

1. Editors, please review the draft, comparing it with the previous one and checking the comments I make above about changes.
2. Consider carefully the need to yield on some points; we all have had to do that. Consider the value of this draft for the Project.
3. Note errors, omissions, or final small fixes in the subsection below, in orderly and considerate brevity.
4. If you absolutely must object to anything, do so soon and clearly, proposing a precise solution, rather than calling vaguely for something to be redrafted.
5. Remember that we have been at this for months, and that we are four days overdue already. Eight weeks (effectively) ought to have been enough time to settle all the issues.
6. Please have your say within 24 hours from now. After 24 hours, I will take whatever next step is appropriate (since I am used to dealing with this, I might as well be the one to carry it through). And we can get a completed version incorporated into WP:MOS.

Endorsements, and final points to fix

[Please sign off on final draft 1; or make final suggestions to fix things]

• (Remaining mute for a moment) My contribution to this has been to assist in making the process work (put a stop to sniping and ensure that a proper consensus once again rules and is immune to tendentiousness). I have not been paying close attention to the details of Noetica’s proposal. Since this is a final call for detail-oriented types to ensure things are correct, I’ll let others go first.

  I invite the first individual here who has a firm grasp of the applicable issues (American practices, other-dialect English-language practices, familiarity with other well respected manuals of style, and the art of compromise) to scrutinize Noetica’s latest version and weigh in here with their thoughts.

  And remember, none of this has been etched by lightning into stone; consensus can change. Furthermore, after this “goes to press,” we can always go back to collegial serial tweaking to improve it. But serial tweaking can’t happen until the admins see first-hand that after unlocking WP:MOS, things don’t degenerate like it’s the chimp pen at the zoo after too-few bananas have been tossed in. Greg L (talk) 14:45, 20 July 2011 (UTC)

• Unacceptable. Does not reflect consensus now; does not reflect a majority; does not reflect the general trend of reliable sources (cherry-picking obscure and rarely used style manuals to support a preconceived position is not following Reliable Sources); does not represent English usage. It is also much more verbose than Kotniski's draft. Yet the changes which would make it tolerable are relatively small:
  • In section 1: the parenthesis should read: (perhaps instead of a semicolon) or be omitted; Another "planet" was detected: but it was later found to be a moon of Saturn is not current English, where a semicolon would be. (I would omit but with the dash as well as the semicolon.)
  • In section 2: Change the first sentence to, for example: There are four ways of forming English compounds: with a space black bird, with no space blackbird, with a hyphen black-bird, and with an en dash. Wikipedia prefers the last for relationships involving parallel, symmetric, equal, oppositional, or at least separate or independent elements; idiom may differ. The examples are far more than necessary, and are often not consensus.
  • In section 3: Change Use this punctuation... to Recast this form where possible; when it is necessary to use it, Wikipedians generally prefer using an en dash in:
  • Remove Post–September 11 anti-war movement; no article title is ever "decided as optimal"; especially on controversial subjects, they're the best we can do for now.

Of these the last section is the most important; the present text would make any claim that this guideline is consensus into a falsehood.Septentrionalis PMAnderson 16:03, 20 July 2011 (UTC)

• The one thing we certainly agree upon, PMA, is that this doesn’t reflect consensus *now*. Of course, that’s only because you are the only one so far who has expressed an opinion. If there are two or three editors who support this and have similar logic, then there will be a consensus, albeit not one to your liking.

  Your unfortunate choice for your first word above (“Unacceptable” rather than “Oppose”) seemed to be chosen to signal intransigence and an unwillingness to budge much in the ‘compromise’ department. At least I hope that’s all it was meant to telegraph, because if you think you are going to use tendentiousness or editwarring if the consensus here is to adopt it in order to impede progress, think again.

  Oh… another thing: Tagbombing WP:MOS after an issue has been discussed and a consensus arrived at, as if those tags are a {{ButIStillWantMyWAAAAAY!}}-tags, will no longer work. {{DISPUTE}}, {{POV}}, {{DISCUSS}} {{WHINE}} tags are not to be used as “Well then, EAT THIS!”Template:Nbhyphgraffiti to force protracted discussion and eventual concessions to the losing side in a debate after a consensus is clear.

  Now I’m gonna go pop some popcorn, pull up a chair, and watch as others resist the temptation to (try to) debate you while they instead merely express their views below so we can discern what the true consensus is… Greg L (talk) 16:30, 20 July 2011 (UTC)

At last, Greg L makes himself clear. We spent a month on a poll, which showed that there is no consensus on what to do here; by Noetica's count, there is the slimmest possible majority. So it doesn't count; only the present discussion among half-a-dozen regulars counts. Business as usual. Septentrionalis PMAnderson 16:54, 20 July 2011 (UTC)

I suppose it takes two to communicate, but you finally got my message point. Yes, the !vote outcomes of past debates don’t factor in to deciding whether a proposal is acceptable now. That’s how compromise and progress is made. Some editors at first have a particular view but later change it after they see tweaks and compromises and buyTemplate:Nbhyphin from others. Some editors fatigue after protracted debate and defer instead to the opinion of other editors they perceive as specialists who are more well informed.

Noetica has labored and repeatedly revised in order to consolidate the best teachings and observations from a variety of editors. What is now under consideration is the newest amalgam designed to achieve a general consensus. It apparently contains elements you opposed in the past, still oppose today, and see as a deal breaker. Understood. Got it.

But I am pleased to hear that you now recognize that it is back to “business as usual.” If that means “WP:MOS and WT:MOS will follow WP:CONSENSUS to the tee,” great. If you think it means “The sword of tendentiousness shall smite those who just talk and agree on things that piss me off,” then, no, it is not back to business as usual.

One thing now that we will be sure to do is run this comment period long enough to ensure that all who are inclined to weigh in here are afforded an opportunity to do so. Greg L (talk) 17:17, 20 July 2011 (UTC)

It contains elements which half those polled opposed. If we are operating by consensus, that is a deal breaker. If we are not operating by consensus, this is not a guideline. Septentrionalis PMAnderson 17:46, 20 July 2011 (UTC)

Support, though not necessarily more than Kotniski's draft. (I'd have to see how that has been edited since I last looked at it.) I was out of the country during the discussion of the first draught on this page, so these are my first comments on it.

• 1. Best absorbed were wavelengths in the range 28 mm – 17 m.

I find this difficult to read, and would spell it out instead. IMO it is not a good example. But IMO we may need an example like "17–28 mm", since there has been at least one editor who's argued that the measurement-unit for the first number is undefined in such examples.

• 2. We need to state up front that we do not extend this convention to bound morphemes like Franco-. (Some MoS's I've seen do en-dash such forms, so this may not be obvious to everybody.)

the Uganda–Tanzania War; the Roman–Syrian War

This is getting repetitive. Do we lose anything by moving one up into the rivalry example?

or even the protein-to-fat ratio

Remove "even". That suggests this is somehow aberrant, when it's perfectly acceptable.

Following the dominant convention, a hyphen is used in compounded place names and language names, not an en dash.

Guinea-Bissau; Austria-Hungary

an old Mon-Khmer dialect; Niger-Congo phonology

Change the rational for the first and remove the second.

(1) Both countries are so punctuated because they are single entities. Compare Congo-Brazzaville and Congo-Kinshasa. Although Guinea-Bissau and Austria-Hungary are good examples (I could see an edit war over whether Austria-Hungary is one entity or two), the reason for them is not arbitrary, as our wording implies. Also, the claim is false: although only hyphens occur at a national level, at a local level (state parks and the like) official names may have a dash, even an em dash.

(2) We do not have a convention of hyphens rather than dashes in language family names. All of the "Amerind" language families, for example, currently use en dashes (following, inter alia, the Cambridge Language Surveys such as The Andean Languages), and there have been no objections. Therefore we do not need to spell out an exception for them.

Also, shouldn't we spell out somewhere our conventions for "hyphenated" ethnicities, rather than one out-of-context example? At least a link?

• 3. Otherwise recasting is better.

Otherwise recasting may be better. It often is. But often the dashed form is more straightforward and easier to process. We should not make a blanket pronouncement when editors may agree that not recasting is better style in some cases.

• 4. Several people have noted that colons are generally acceptable here. I agree; I've come across several list articles so chock-full of dashes that they've become difficult to read. Colons often look better, though in some formats they may not be appropriate. We should provide them as an explicit alternative.
• Per the section below, while carbon–carbon bond should use an en dash, C−C should probably use a minus sign, parallel to C=C and C≡C, as the arithmetical signs of a good font are designed to be the same width and centered at the same height, but do not in general correlate to dashes. Maybe a note?

— kwami (talk) 18:08, 20 July 2011 (UTC)

• Support From what I can see, this effort to simplify is a magnificent effort by an Noetica to distill something around which a consensus can form. Any time there is a shepherding author like that, he or she is über familiar with the details to an extent I can never be. But it seems too that others here, including User:kwami, have pored over it as best they can and, while seeing some items that could be tweaked, are overall supportive in the spirit of compromise and getting back to a sense of normalcy. That basically describes my attitude. It may not be perfect, but it is a portion (dealing with dashes), of a manual of style at an all-volunteer hobby site. No condemned prisoners will be executed this week over what is being decided here. Greg L (talk) 02:41, 21 July 2011 (UTC)
• Support – as I read it, Noetica has done as good a job as possible at incorporating all the comments into a new version that's cleaner, and closer to true consensus, than anything we've seen before. None of these provisions will be cast in stone, but we'll be able to unlock things and get back to a process of orderly change by discussion of individual points, instead of going back to the edit wars over such things as the "optionally" in the "ex–prime minister"; it doesn't really matter at this point how particularly difficult points like this get decided, except that the decisions at least have majority support, which they now all do, I'm pretty sure. We can fine tune it after this is done; what we can't do is keep the discussions open on multiple fronts at once, while keeping the MOS locked, and while PMA obstructs all efforts to get it settled. Dicklyon (talk) 04:59, 21 July 2011 (UTC)
• Support, and if I don't post in this section again within 48 hours, I'll have nothing in detail to say (I'm on a ruinously expensive timed connection for the next day and a half). Tony (talk) 06:04, 21 July 2011 (UTC)
• Oppose, for reasons previously stated about spacing. The objection is partially procedural (I simply cannot see a “clear majority”), and it would appear that no consideration was given to other than “I like it”. I hate to say it yet again, but the guides agree with me. At least two usages introduce needless ambiguity, and I regard this as more serious than others such as whether the dash is spaced in Christmas Day – New Year’s Eve. Though it is not a basis of objection, I again strongly recommend that nonbreaking spaces be included in the examples. The examples themselves aren’t likely to break, but editors (who might examine the code) learn by example. JeffConrad (talk) 06:53, 21 July 2011 (UTC)
• Support. a clear explanation of a very complex subject. Noetica has been scrupulous in taking note of the various points made in the discussion. Colonies Chris (talk) 09:23, 21 July 2011 (UTC)
• Support. I applaud the good work of Noetica, and would say the time for filibustering is over, bearing in mind the lengthy discussions and overwhelming endorsement by the community. There may be some small details to work out, which would be a lot easier dealt with as amendments to this once passed. 88.160.245.226 (talk) (aka Ohconfucius) 10:05, 21 July 2011 (UTC)
• Support. I'm less concerned about the issue than most people here and I'm unlikely to be active in implementation. However, too many experienced editors have wasted too much time under the current arrangements. I'm happy that there is a workable proposal on the table and I think it's time to accept it. Lightmouse (talk) 16:28, 22 July 2011 (UTC)
• Support. As per Greg L. GFHandel ♬  00:21, 23 July 2011 (UTC)

Interim replies (after 24 hours)

I respond here to all who have posted so far in the subsection above: #Endorsements, and final points to fix. I propose now to wait a further 24 hours for any new commenters to contribute their endorsements or final fixes. New commenters: please post in the next subsection: #Endorsements, and final points to fix (II), leaving the first one devoted to issues that others have already raised above. Please try to keep things orderly, as we work through the last phase of this protracted dialogue. Noetica^Tea? 13:00, 21 July 2011 (UTC)

Answering Greg

The experience you bring to the table, along with your continued attention to details of process, is invaluable. I hope you will stay to watch developments, and comment as you see fit as we step methodically toward a resolution here. Just one quibble: your observations on consensus apply generally, and of course the policy provisions at WP:CONSENSUS are not to be set aside. But I remind you, and all editors here: we work under an Arbcom injunction, and with detailed instructions from ArbCom member Casliber. For a summary of that rather unusual background, editors can look at the beginning an earlier section of this page. Consensus was supposed to be assessed by ArbCom on 16 July 2011; so we are five days behind already (effectively after eight weeks of deliberation and then voting). The present task is how to reflect the consensus established on the voting subpage, and associated discussion, for ArbCom to evaluate. That does not mean we fall silent on remaining points of contention; but it might show them in a different light.

Noetica^Tea? 13:00, 21 July 2011 (UTC)

Answering PMAnderson

Your first comment above is headed "Unacceptable". I think no one is surprised at this, and nothing I could in good faith say to you would make an iota of difference. I will address matters of fact on which we disagree, and also extract what I can that does address the purpose of the present phase of discussion.

• You write:

Does not reflect consensus now; does not reflect a majority; does not reflect the general trend of reliable sources (cherry-picking obscure and rarely used style manuals to support a preconceived position is not following Reliable Sources); does not represent English usage.

I answer that the draft reflects consensus better anything so far offered, and in fact is not distant from the dash guidelines it is intended to replace – which were also the result of careful collegial work. The draft does reflect a majority (a point I return to below). It most certainly reflects the trend of reliable sources, which are in this case recent general and specialised style guides, major dictionaries (including OED, against which I checked several examples), and related sources. It does reflect best-practice contemporary publishing, as codified in the sources just mentioned. Far from anything in the process being biased toward some "preconceived position", there was perfectly open process and the widest consultation, with anyone able to adduce any source they wanted to. If the present draft "cherry-picked" at all, it selected from palpably weighty opinions in the voting, and as little as possible – only as necessary to secure the nearest to a perfect consensus that we can achieve in an imperfect world, especially in this stubbornly refractory corner of the real world.

• You continue:

Yet the changes which would make it tolerable are relatively small: ...

I answer that those changes are not timely (you had the last week to raise them), not consensual (not voted on, discussed, or otherwise tested), not focused on dashes (though you yourself set up the subpage with a limitation to discussion of dashes, against advice that I put on record), not designed to lessen prolonged and life-wasting wrangling at thousands of content talkpages, not specific as I had requested at the head of the section in which you comment (you refer to examples, but do not say which), and not all "relatively small" compared to the changes others propose here. More could be said; but I confine myself to your objection to "optimal". That one is relatively small; but I find nothing useful in it, since what is optimal in any case clearly amounts to "the best we can do for now", as you put it. Revision is always possible, or course. No one thinks otherwise. Nevertheless:

• You write concerning that last point on optimality:

Of these the last section is the most important; the present text would make any claim that this guideline is consensus into a falsehood.

I don't agree; but to remove what you take to be "most important" objection, I will amend the wording in the next draft I present.

• I reproduce your later post in full:

At last, Greg L makes himself clear. We spent a month on a poll, which showed that there is no consensus on what to do here; by Noetica's count, there is the slimmest possible majority. So it doesn't count; only the present discussion among half-a-dozen regulars counts. Business as usual.

Greg always makes himself clear. It's refreshing! Some corrections of fact, though:

• No, we spent more than six weeks on a poll; 60 editors contributed to that page, and it showed dramatic agreement with almost all provisions of the existing dash guidelines. Where there was diversity of opinion, analysis still shows that a majority supported the guideline in question. Useful discussion ensued, and we can draw on it here to make an even more robust, comprehensive, and consensual version of those guidelines.
• No, again you misrepresent what I have said. Only by applying a severely prejudicial spin can the hard quantitative facts be read as you suggest. You claim that mere voting is not enough. I agree: we must analyse, and look at the reasons and codicils. I have done so accurately; you have not. It was possible to find an equitable solution in the case of every guideline. Not easy, but possible. The results are in the present draft, soon to be refined even further. In any case, your recently discovered cynosure, WP:CONSENSUS, guides all our actions here. And it provides for majority determination where all else fails. You seem to be doing your best to ensure that all else fails. Accept the consequences.
• No, not half a dozen regulars. This time you find yourself opposed by the majority of those who voted, in the very subpage that you set up.

I will pass over in silence your later post. It adds nothing that I can constructively address.

Noetica^Tea? 13:00, 21 July 2011 (UTC)

By Noetica's count, I find myself opposed by 15 out of 29 editors. I don't think that count is correct, but that is not the point at issue; if it were correct, it would not be consensus.

This is not a technicality: if there is genuine consensus in this matter, there would be reason to suppose that most editors not surveyed would agree with the consensus; and we could continue on that ground until shown wrong. But there is no real reason to suppose that so slim a majority represents even a majority of editors; under our uncontrolled conditions, it may or may not, as likely the one as the other. Septentrionalis PMAnderson 20:11, 21 July 2011 (UTC)

Noetica's claim of untimeliness is false in fact; I did object last week. It is, more importantly, another policy violation: Noetica has invented rules without consultation, and without stating them, which would require anybody who objects to this proposal to log on in a certain time. But participation in Wikipedia is not compulsory. That is policy. Septentrionalis PMAnderson 20:32, 21 July 2011 (UTC)

Follow-up answers for PMAnderson

• You write: "By Noetica's count, I find myself opposed by 15 out of 29 editors." In doing so, you once again distort the truth so transparently that I can comfortably defer the refutation while I deal instead with serious talk, elsewhere in the section
• Since that palpable misrepresentation is the premise for your next point, I do not dignify it with a separate response.
• If you think me guilty of "policy violations", or single me out (!) as trifling irresponsibly with Wikipedian principles or with ArbCom's specific instructions, I sincerely hope you will take the opportunity to expose all of those issues when the imminent final draft is presented to ArbCom for certification, a week late. Who knows what other issues might be brought forward – by whom, and concerning whom?

Noetica^Tea? 15:02, 22 July 2011 (UTC)

Answering Kwami

Welcome back. A pity that circumstances prevented your coming in earlier. Thanks for the points you now offer. My answers:

• On "wavelengths in the range 28 mm – 17 m", I agree that this form is not the easiest to read. The use of en dash is not the only option, of course; but it seemed to me from discussion that this was considered the best implementation, if one had already settled on an en dash. (I return to this below.)
• On "bound morphemes" like Franco~, I had thought the point was solidly made in the examples. There are tricky cases to add (Sino~North Korean détente??). I have something to say about this, but I hold off for now. Let's deal with it in the normal course of discussion, once all this business is behind us. It's way past everyone's bedtime!
• You mention these two examples: "the Uganda–Tanzania War; the Roman–Syrian War", and you comment:

This is getting repetitive. Do we lose anything by moving one up into the rivalry example?

I smiled a wry smile when I read that. I simply invite you to reflect on the prehistory of the present dialogue, involving a certain North American war. Yes: we need both. (Remember also your own subtle alteration of the text on the voting page, to include an {adj–adj noun} case.)

• You want even omitted from "or even the protein-to-fat ratio"? Consider it done!
• You comment on "Following the dominant convention, a hyphen is used in compounded place names and language names, not an en dash." You say several things; I pick up first on this part:

Although Guinea-Bissau and Austria-Hungary are good examples (I could see an edit war over whether Austria-Hungary is one entity or two), the reason for them is not arbitrary, as our wording implies. Also, the claim is false: although only hyphens occur at a national level, at a local level (state parks and the like) official names may have a dash, even an em dash.

I do not think our wording suggests that the matter is arbitrary; it simply appeals to a "dominant convention". The claim concerning its dominance is not false: it dominates in the sense that it has strong adherence, statistically and in regimented practice. What did you read "dominant" to mean?

• You write:

We do not have a convention of hyphens rather than dashes in language family names.

OK. But we need one; and the use of hyphens is by far the dominant convention here. In this case and also for countries, why not endorse what the major sources almost universally choose to do, in a painfully thorny area? You can cite exceptions; so can I. But to what end? I say we want a guideline; and I submit that we don't want a patchwork of disparate, perpetually disputed titles for articles on kindred themes.

• On recasting cases like ex–prime minister Thatcher, you want a softening. Others want a hardening, so that such forms are always recast. Still others (like myself) detest that form, and would want a hyphen or recasting, and never an en dash. The draft form of this guideline is a cunningly crafted compromise. No one is completely happy; but everyone should be able to live with it. The pain is shared, and the guideline serves to defuse hundreds of future disputes – if people will consult MOS for a recommendation based on sound principles and precedents.
• On colons and their preferred use, I agree. But the point is already made, isn't it? I suggest you read the draft guideline for sentence-level dashes again.
• On the carbon–carbon bond: the guidelines address continuous prose, not specialised chemical notations. We have WP:MOSCHEM for that. But we can't point to every location that may be peripherally relevant. The guidelines must be readable and focused, and show carefully selected indicative examples.
• On your extended remarks concerning hyphens and slashes, nothing needs to be done here. For eight weeks we have discussed dashes, and hyphens and slashes only as they bear most directly on questions concerning en dashes. Another time! When the present business is finished.

Kwami, I hope you will continue to contribute in these final days of the process. Your expertise is much appreciated. Please remember: these are the final, polishing stages.

Noetica^Tea? 13:00, 21 July 2011 (UTC)

Okay, most of this is pretty minor. (Of course I remember the Mexican–American War dispute. But I thought we could consolidate those examples so we're not so repetitive.) But re. languages: What you propose is making a specific counter-convention for language families. No other field is made to conflict with the semantic use of dashes, not even geographic names. Why this one? If it's only because most sources use hyphens, as you say, that was Anderson's argument for abolishing dashes altogether, so again, why this one? I don't see why all of the language-family articles currently using en dashes should be changed to hyphens just because ... what was the reason again?

As for country names, please consider rewording from

Following the dominant convention, a hyphen is used in compounded place names

to

A hyphen is used in compounded (self-standing?*) country names

Otherwise we're going to have problems when official place names have dashes, for example Ed Bird – Estella Lakes Provincial Park as punctuated in the Protected Areas of British Columbia Act, and people on those articles decide that the MOS is therefore useless for deciding punctuation. *I also worry about cases like the Polish–Lithuanian Commonwealth. Perhaps that example could be added to the 'border' and 'treaty' examples? — kwami (talk) 19:32, 21 July 2011 (UTC)

It would be nice if discussions of my position were based on something I had actually said. I do not propose to "abolish dashes altogether"; I never have. I support permitting hyphens in words where they are normal usage. Septentrionalis PMAnderson 21:35, 21 July 2011 (UTC)

Thanks for telling us. I did not understand that from your arguments up to now. — kwami (talk) 00:20, 22 July 2011 (UTC)

Sort of like he fought to "permit" a hyphen in Mexican-American War, even though it was long stable with an en dash. Dicklyon (talk) 00:39, 22 July 2011 (UTC)

Well, yes, this would include permission to spell Mexican-American War the way the sources do; whether we use a permission in any given case is a matter of judgment (in which I am not alone. Septentrionalis PMAnderson 01:44, 22 July 2011 (UTC)

Follow-up answers for Kwami

I have thought long and hard about your suggestions, and done some delving. I have changed my mind about one thing.

• I still think that the examples are not viciously repetitive. They make subtle distinctions despite appearances; and we know from experience what vexations eliding such distinctions can lead to. I am not inclined to alter the set of examples generally.
• On the other hand, I am ready to remove the entire guideline for compound place-names and languages (including language families). I think there will be no serious dissent. There does seem to be a "dominant convention" in favour of hyphens, looming over other considerations (cf. chemistry's local customs, or the post-Linnaean niceties of botany and zoology); but I concede that it is not overwhelming. You mention The Andean Languages; well, Cambridge does indeed use en dashes and hyphens to make distinctions that others obliterate. Routledge follows very similar conventions; and over the last year I have been collecting examples from Glanville Price's Encyclopedia of the Languages of Europe (Blackwell, 1998) that must cheer the pedant's flinty heart: "Basque–Icelandic pidgin", but "bilingual Cornish/English speakers"; "Abkhaz-Abaza" (Abaza having two dialects: Bzyp and Abzhui; who knew?☺), but "a single Abkhaz–Abaza dialect continuum" all the way from Abzhywa to Ta'ap'anta; "Chechen–Ingush Autonomous Soviet Socialist Republic" and "a joint Chechen–Ingush written language", but "the Chechen-Ingush speech community" (I seem to recall they had a war about en dashes). The four-volume International Encyclopedia of Linguistics (Oxford; I only have the superseded 1992 edition) uses hyphens throughout ("Niger-Congo languages", as against Oxford's core style references and OED: "Niger–Congo languages"), but en dash in cases like "New Ireland–Tolai languages" and "North Trans–New Guinea Languages". In short, in these important reference works semantically based consistency is the principle, but managed and attenuated according to some general standard (Cambridge or quasi-CMOS, in the cases I have examined). Semantic distinctions are extraordinarily refined in this domain, and many standard works lapse into frank inconsistency, like Brown and Ogilvie's Concise Encyclopedia of Languages of the World (Elsevier, 2008), which alternates between "Uto–Aztecan" and "Uto-Aztecan". Let them work it out among themselves, and let's defer the issue for special consideration later. Our general guidelines will suffice till then. As I have suggested, the same should apply to the other matters you raise (so late, perforce); they can and should be addressed when we come to review the guidelines for hyphens and slashes, or some other opportunity.

Noetica^Tea? 13:27, 22 July 2011 (UTC)

I was going to suggest removing those myself. If there is evidence of an exception needed in some field, the relevant project can take that up on their own style page. In looking at books for en dashes in compound language family names, I did find a few, so I suspect there's no real rule about that one way or the other. Dicklyon (talk) 13:55, 22 July 2011 (UTC)

How bout we leave in the country examples? Guinea-Bissau is a dab from Guinea-Conakry (though no-one uses the latter), so these are not co-equal elements. But it could easily cause confusion, since a country and its capital are two independent things. Austria-Hungary is an excellent example, since you really could debate that one ad nauseum. And I think Polish–Lithuanian Commonwealth, or something like it, would be good to include, so that people don't become absolutist. I agree with leaving the language families out; they can be addressed, as you say, at the wikiproject if they are disputed. — kwami (talk) 17:07, 22 July 2011 (UTC)

How would you debate Austria-Hungary? Who uses anything but a hyphen? Who recommends anything else? Septentrionalis PMAnderson 20:31, 22 July 2011 (UTC)

Answering Dicklyon

Thank you for respecting the intention of this phase of the process. Yes, of course we can fine-tune things later. Consensus can change, and the guidelines can be refined, and polished to reflect consensus even better.

Noetica^Tea? 13:00, 21 July 2011 (UTC)

Answering Tony

Your terse support is helpful as we advance toward a resolution. But do get back to a proper connection soon!

Noetica^Tea? 13:00, 21 July 2011 (UTC)

Answering Jeff Conrad

I don't know what others think Jeff, but I acknowledge your obvious experience and expertise. I know that you don't like everything in the present draft. As I have pointed out, I don't like it all either! In fact, I would probably object to more in it that you. Please think again about that. Neither of us is writing a guide to epitomise his own practice. On specific points:

• You write:

The objection is partially procedural (I simply cannot see a "clear majority"), and it would appear that no consideration was given to other than "I like it".

Well, you at first objected (almost a week ago) that a "clear majority" was not something to be looking for. Well, what is to determine the matters that you contest? What is your analysis of the numbers, and the comments, and the ensuing discussion, on those matters? And do you suggest that I act upon the principle "I like it"? Where, in the name of all that's Jimbic? I have deliberately set aside what I like. (Have you, by the way? As firmly as I have?) I am assuming that we both want guidelines that will help prevent the month-long wars over article titles that led to all of this, right? Well, at the head of the section in which you comment, I made this request:

If you absolutely must object to anything, do so soon and clearly, proposing a precise solution, rather than calling vaguely for something to be redrafted.

That, I suggest, was a fair request; but you have not done what I fairly asked. So now, here is my specific request of you; please attend to it urgently:

Jeff, please draft and show here your own redrafting of the guideline for ranges with spaced components: in full, with examples. Do not shy way from controversial cases that might come up at several talkpages of articles, as the present draft does not. In short, make a guideline that guides, and omits nothing relevant. Then, please, do the same for the guideline concerning ex–prime minister Thatcher and such cases. Complete, and easing the problem of replicated disputes at as many talkpages as possible.

That is no playful or petulant request. I'm serious. If you can capture consensus better and with less likelihood of ruinous disputes arising, your alternatives should be considered.

• On Christmas Day – New Year’s Eve, I have twice answered your objections and explained why it is a date range, how it differs semantically from New York–Chicago flight. Please address that case in your draft guideline; and please also show how you would punctuate Christmas 2001[~]Easter 2002, and include the principles by which you determine these things.
• You write:

I again strongly recommend that nonbreaking spaces be included in the examples.

I understand why; and I have answered that. You have not answered me. You had a week to raise the matter, but did not until extremely late, delaying my scheduled posting (of which I gave 24 hours' notice). No one else has called for this, and there was agreement earlier not to complicate the dash guidelines with matters that are more efficiently treated generally. The general treatment could show en dashes with &nbsp;, but there is no call for cluttering the dash guidelines in this way. They are intricate enough!

• I invite you to reflect on the enormous difficulty of the present task. I know you have addressed it energetically. Please consider, though, how long all of this has taken me. How much work lies ahead depends partly on you, now. Please take some questions that I put above as rhetorical; shorten discussion, and make life easier. I have done all I can to meet your concerns. Now meet mine! Are you certain that holding out is worthwhile, where I and others suppress objections that are at least comparable to your own?

Noetica^Tea? 13:00, 21 July 2011 (UTC)

A few comments. I’m sorry for the length, but don’t know how to cover the issues in a couple of (silent) sound bites—that’s why I simply referred to the guides in the polling. I hope this answers some of your questions.

---

---

• “I like it” was directed at the votes under 6b. when there is a space within either one or both of the items, not at you—I’m sorry if I did not make this clear (musta been the typographical quotes . . ). As nearly as I can tell, the only ones to state much of anything other than “I (don’t) like it” were me and macwhiz, who cited style guides—I cited many, macwhiz cited CMOS16. In a sense, I could tally the voting as 2–0 in favor of closed-up usage, but that would obviously be silly. The issue that I see as having been given little of any consideration is a simple one: given that most of the high-ASR (Amazon Sales Rank) guides call for closed-up use, why would we diverge? You mentioned five major guides, but did not indicate what they were, so let me begin with what I was referring to. A while back, we looked at ASRs of several widely used guides on Amazon’s US and UK sites. I repeated this exercise a couple of days ago, adding a couple of glaring omissions, especially Butcher, and included Amazon Canada (they don’t seem to have one for Australia). I eliminated guides that did not deal with en dashes from the final results, and was left with 16 guides. I can provide the complete list of results if needed, but I’ll summarize for now:
  • US:Of the 16 guides, all but #10 (CGEL) and #16 (BCE) call for closed-up use exclusively.
  • Canada:All but #11 (BCE) and #12 (CGEL) call for closed-up use.
  • UK:All but #5 (BCE) and #9 (CGEL) call for close-up use.

I don’t have CGEL, so I don’t know what it calls for. And full disclosure: the results include CMOS 14–16 and Turabian, so there is arguably a bias. Nonetheless, the older versions of CMOS sell far better than I would have guessed.

We’ve both acknowledged the many limitations of ASR, but absent access to expensive commercial publishing databases, ASR seems better than simply “top guides” (the ones I like, of course). Though it might be nice to cite “de facto” practice in some cases, I think doing so in a verifiable and statistically valid way would be nearly impossible. So perhaps ASR could be regarded simply as a SWAG rather than a mere WAG. Accordingly, I don’t suggest that the recommendations of the listed guides are necessarily dispositive, but I do suggest that if we are to diverge from what seems to be fairly broad practice, there should be at least some reason. So much has been said here that it’s tough to track them all, but I don’t recall recent mention of why we should space en dashes at all. The only guide that I’ve examined that calls for spacing is BCE, so I’ll start with that:

However, spaced en rules may be spaced between groups of numbers and words to avoid implying a closer relationship between the words or numbers next to the en rule than between each of these and the rest of its group: [emphasis added; search for some of this text above for the examples]

The assumption is apparently that the en dash binds more tightly than the space; although this hardly seems unreasonable, it clearly is not universally held, especially in North America (and at least nominally, at OUP). As I have said, it’s tough to make the case that one practice is “right” and that the other is “wrong”, but that the preponderance of the guides, especially if weighted for ASR, call for closed-up use suggests to me that blowing off NA practice should at least have some basis.

As I’ve also said, one great advantage of NA practice is that there is little likelihood of confusion among the em dash, en dash, and minus. Forcing the em dash is clearly a nonstarter, because that’s just not the way many British publishers do it, and they aren’t likely to change. But although it’s not my preference, there is little likelihood of confusion with a minus, or with a range/whatever en dash if the latter is closed up.

I’ve suggested that the issue is one of “operator precedence”. At least in NA, it would seem to be something like “.”, “-”, “ ”, and “–”; with such an order, no ambiguity arises from closed-up use. With much British practice, the perceived order of the space and en dash would seem to be reversed, and some ambiguity could conceivably result. Practically, I agree with A. di M. that anyone in her right mind would fairly quickly recognize that “28 mm – 17 m” meant a range rather than a difference, and the same is probably true for “20 kHz – 20 Hz”. But let’s face it, the same is true for “28 mm–17 m” and “20 kHz–20 Hz”; with either approach, there may be some “initial ambiguity” that slightly interrupts the flow of reading, but in most cases, the meaning is probably quickly resolved. In this case, I would maintain that the “initial ambiguity” arising from the spaced use is considerably greater given the strong association of a spaced minus-kinda thingie with subtraction. Perhaps it’s just be, but with CSE and NIST I seem to be in reasonable company. And, whether they had this specific concern in mind, the significant number of general-purpose guides that recommend closed-up use arrive at the same end. Incidentally, my reference to “not very nice people” above was not meant to imply that that “the 20 kHz–20 Hz propagation delay” was a not very nice usage. But if Frankfurter had ever had to deal with en dashes, he well might have chosen similar words . . ..

As several of us have stated, “28  to 17 m” well may the best approach when used as a nominal, but with adjectival use, the close-up dash seems the least of evils.

• Ranges of dates. Much the same applies to ranges of dates, although at least to me, the initial impression of subtraction is slightly less than with physical quantities because I don’t normally subtract dates. Nonetheless, I maintain that there is no greater “initial ambiguity” with “21 July–16 August” than with “21 July – 16 August”. And I return to the example of “21 July – 3 September 2012”: is it the period between now and then, or just the duration of the Obama–Palin debates? Here is a case of genuine ambiguity rather than simply “initial ambiguity”. Using “21 July–3 September 2012” avoids the genuine ambiguity. From this, it should be clear how I would handle “Christmas 2001–Easter 2002”. And the rule is simple with NA and OUP practice—always close up. A fallback rule might be inferred from BCE: space when there is a number on one side of the dash and alphabetics on the other. But I would look more to her use of may, because the preponderance of bestselling guides don’t seem to find ambiguity with close-up use. Again, I don’t suggest that there is only one way to do it. But if indeed we generally follow the sources, the advice of a seeming preponderance of guides would seem to merit serious consideration, which I don’t think it was given.

Now perhaps my explanation is eleventh hour. But my citation of a plethora of guides was not, and their recommendations seem to overwhelmingly support the position I advocated. As I mentioned, I would be OK with spacing the dash in complete ranges, especially because, as Tony mentioned, there are zillions of these at the beginnings of biographies, and I think it better to follow the well-established style in that instance. In most other cases, I would probably use to or through.

• Nonbreaking spaces.' My intent here was only to include them in the code for the examples, not to attempt to explain them—it’s covered elsewhere, as you noted. But if they’re missing from the examples, it would seem that the MOS would fail to follow its own advice. And the MOS is the potent, the omnipresent teacher. For good or for ill, it teaches all editors by its example. JeffConrad (talk) 23:16, 21 July 2011 (UTC)

I’ve added nonbreaking spaces to a few dates (and the quantity range) in the draft. I did not add the space to 1–17 September; I would probably use one (it’s no worse than 10:30 pm), but wonder what others think. Whether or not the space is added, this clearly is an issue to be covered elsewhere. I′ve avoided any of the other changes discussed above on the assumption they would not be well received. JeffConrad (talk) 04:36, 22 July 2011 (UTC)

Follow-up answers for Jeff Conrad

• Thanks for adding &nbsp; in the code for the draft. I was deferring that until it was finalised, but it's good that it's done now. (For 1–17 September I would consider context; I think it's marginal.)
• For some guides that agree pretty well with Butcher's on spacing of en dashes, see here in the archives of this talkpage (search on "These sources may be of some value").
• I had not wanted to raise it, since I wanted the discussion of spacing in ranges to be fresh and unprejudiced, but in fact WP:MOS already calls elsewhere for spacing such as most voters favour in our recent poll. See Wikipedia:Manual_of_Style#Chronological_items, and search on "range". For more detail, and for treatment of other ranges, see Wikipedia:Manual_of_Style_(dates_and_numbers)#Unit_names_and_symbols (for example) and again search on "range".
• I am as ready as anyone else to overturn anything irrational or that lacks consensus; but those guidelines are rational and consensual, as established in long deliberations at two talkpages with many participants (search the archives). The only things that are striking in the present draft are two examples: Christmas Day – New Year's Eve and Christmas 2001 – Easter 2002. They accord perfectly with the principle laid out in the draft, and with the principles already used in WP:MOS and WP:MOSNUM.
• Do you still want to make a different principle (or principles) for those two date cases and for the number ranges with changed units? I have twice asked you, very explicitly, to draft what you would prefer. You have not done so, but continued to make points generally without showing any black-letter proposal. What do you expect me to do? How long would you like me to second-guess what hard-and-fast wording would satisfy your need?
• I could very easily roll over and remove something from the draft or soften something else – or make something vague, which more than one style guide has done when faced with an impasse. I for one have no time left to go again over the details that you and your interlocutors produced over weeks of discussion. You had every opportunity to tease out salient issues and resolve them, or to push the boundaries and find the interesting cases that I have adduced and put in the draft. But you did not. The principles tested in voting and discussion on the subpages are further tested in the hard examples I produce. Do they stand up? I think so! If you disagree, there must be something wrong with the principles you and others forged or agreed upon, not with the examples. I have asked you to show otherwise; you have not.
• I can with equal ease take stock right now, and deliver a consensus to ArbCom to certify. The process has been long, exhaustive, and exhausting. Everyone has compromised; and who can find anything new to say? I can't, beyond pointing to long-established guidelines here that appear to be confirmed in fresh examination by many editors.
• On the grounds laid out above, and earlier, I make a last appeal to you: concede a little more. We all have (with the exception of PMAnderson, of course). I value your knowledge and acumen; and your imprimatur would give a further seal to a draft that already does all that anyone could have hoped for.

After another 24 hours I want to lock in a truly final draft, and take the next steps for its incorporation in WP:MOS. I hope you will support this move – just as almost everyone else has, through collegial discussion and negotiated concessions. You will understand if I cannot wait any longer though, nor enter into still more wordy to-and-fro. Eight weeks (with months before that, of bickering) is about my limit. And I know I speak for others too when I say that.

Noetica^Tea? 14:46, 22 July 2011 (UTC)

You need merely change as suggested, or at least in that direction. That would settle the bickering; that is what consensus is supposed to do.

I applaud JeffConrad's method; I would like to say that before this is settled. It shows that on this point, he has the strength of the arguments and the bulk of the style guides with him; in short, he follows the sources. Septentrionalis PMAnderson 20:29, 22 July 2011 (UTC)

• Guides that support Butcher’s advice: Texas State University’s editorial style guide—can I see what the local community college recommends? Can I include a paper that I wrote on this topic over 30 years ago? This isn’t to knock Texas State, but to put it with the more widely used guide suggests desperation. For popularity, The Cambridge Guide to English Usage is in the mid-to-high 400,000s in ASR for Canada, the UK, and the US. If today’s rankings were inserted into the lists I recently compiled, this guide would rank 17, 10, and 10, respectively. The Australia-specific guides barely hit the radar in any of these rankings, but this obviously is misleading because they probably sell far better in Australia. Once again, I acknowledge that only so much can be inferred from ASR, but so far, it’s the best thing I’ve seen proposed here as a reasonable means of quantifying “top” style guides. And the data still suggest that the current and proposed wording is at odds with the overwhelming majority of guides. My objection remains that this wasn’t even considered, and remind that I mentioned this when I voted on 5 June—not a single person responded.
• Discussion of spacing in ranges: I’m willing to stand corrected, but I don’t see much in the discussion about why some range dashes should be spaced—it was largely “I (don’t) like it”. As mentioned, the only thing other than “I (don’t) like it” was the mention of style guides by two of us. Why did I not raise some of these specifics earlier? Recall that the ArbCom motion was to settle a dispute relating to article titles. This polling began as an informal effort by PMAnderson to see where we agree and where we disagree. It apparently morphed into a vote on what the MOS should say about dashes, but without any formal announcement of such. Even with a formal announcement, the procedure that I usually seen almost everywhere is to summarize results, securing agreement that there is agreement with the inferred conclusions, and then moving to the next step. That wasn’t done here, and I still maintain that consensus was not found on at least a couple of items.
• What’s rational and consensual is sometimes in the eye of the beholder. If I were to summarize the archived discussion linked above, I would say that 1), there clearly was no consensus, and 2), the support for the current wording was primarily from two editors who pushed their preference pretty hard. There’s certainly nothing wrong with pushing one’s preferences—I do it just like anyone else—but a hard push doesn’t necessarily indicate consensus. I don’t suggest that the proposed practice is irrational, but simply that it’s no more rational than North American practice—and the latter avoids several problems, so in some cases, the current and proposed practice is irrational. I also suggest, based on the data presented, that it’s the minority practice—this doesn’t mean that it’s wrong, but perhaps does suggest that proscribing the more common practice is not reasonable.

I would finish my comment here by saying that I see closing up range/etc. dashes as affording no disadvantages whatsoever in comparison with spaced usage in some cases, and it does afford a few advantages—closed-up usage largely precludes confusion with parenthetical en dashes and minus signs, and it avoids ambiguity in the instances of two dates followed by the year of the ending date. So at worst, it does no harm, which is essentially what I said to PMAnderson and a few others in cases where the en dash was seen as having little value. To me, that makes North American practice clearly more rational.

• I’m surprised that you think I’ve failed to respond to your requests to say how I would handle it, because I clearly have so indicated:
  1. I would follow majority practice and always use range/etc. en dashes closed up, avoiding the need for any special rules.
  2. Barring that, one possibility would be to draw literally from Butcher’s examples and space range/etc. en dashes when there is a number on one side of the dash and an alphabetic on the other side. I thought the meaning was obvious, but examples would be all date examples in the proposed draft but Christmas Day–New Year’s Eve.
  3. Without additional qualification, dashes would be spaced in ranges of physical quantities with changed (or simply repeated) units. I think I made it clear that, to avoid possible confusion with subtraction, I would not space a dash in a range involving physical quantities, so no special rule is needed for changed simply repeated units, leading to 28 mm–17 m and17 mm–35 mm. Clearly, not everyone seems to see the problem of confusion with a minus (e.g., the EU guide apparently does not), but I do find support from some pretty solid sources in NIST and CSE. I don’t see any real drawback to closed-up usage here, and it does seem to preclude initial confusion with subtraction. Again, at worst it does no harm.
  4. And yet another possibility would be to borrow further from Butcher: en dashes may be spaced in these situations, which in effect would be to allow either North American or non-OUP British practice, much as we treat (and propose to treat) parenthetical dashes. I’ve stated this as well.

I haven’t bothered with formal draft of these points because I think I’ve made it quite clear what I would propose, and without reason to believe that such a draft would get serious consideration, I don’t wish to spend the time or effort. I’ve already invested more time and effort just in this discussion than I’ve otherwise spent on punctuation in my entire life (aside, perhaps, from a similar discussion about quotation marks).

• I hate to be the one in opposition, but I simply cannot join an “across the board consensus” that I cannot see. One possible approach would be to recognize that we don’t really have consensus on this issue (and perhaps on using a dash in open compounds not covered by other rules, e.g., Chuck Berry–style lyrics), and refer the matter for further discussion, as Dick Lyon suggested. The issue of spacing dashes in ranges would have little impact on article titles, and although a title conceivably could be affected by the second issue, actual instances would seem few and far between. I would endorse such an approach.

JeffConrad (talk) 08:45, 23 July 2011 (UTC)

Final responses to Jeff Conrad

Jeff, of course I am disappointed that you could not compromise enough to accept the present draft (soon to be posted with amendments), but of course I respect your decision. Again, there are too many words for this latest of late stages. I will respond to just a little of what you write this time.

• "Guides that support Butcher’s advice: Texas State University’s editorial style guide—can I see what the local community college recommends? Can I include a paper that I wrote on this topic over 30 years ago? This isn’t to knock Texas State, but to put it with the more widely used guide suggests desperation."

That is a misleading and unworthy comment, Jeff. I linked to an archive to keep things simple, and to spare myself yet more typing of text to no effect. I listed six sources, and of this one I wrote: "Texas State University's editorial style guide [...] This is one of several academic sources online that prefer the general style given in sources above, though perhaps implicitly." Do you want still more of the same? Please go looking, as I did.

• "For popularity, The Cambridge Guide to English Usage is in the mid-to-high 400,000s in ASR for Canada, the UK, and the US. If today’s rankings were inserted into the lists I recently compiled, this guide would rank 17, 10, and 10, respectively. The Australia-specific guides barely hit the radar in any of these rankings, but this obviously is misleading because they probably sell far better in Australia. ..."

You go on to acknowledge the limitations of that sort of count. Well you might. We have been over this ground before, and much more that you seem intent on traversing yet again. You don't like "clear majorities", it seems; but you appeal to the numbers when it suits your purpose.

• "My objection remains that this wasn’t even considered, and remind that I mentioned this when I voted on 5 June—not a single person responded."

You should perhaps have insisted at the time; and not diluted a core message with a torrent of text. I took up points of yours at great length, and with meticulous concern for detail. But I can do only so much.

• "Discussion of spacing in ranges: I’m willing to stand corrected, but I don’t see much in the discussion about why some range dashes should be spaced—it was largely 'I (don’t) like it'."

Yes, that is regrettably the way of things, often. Recall that Casliber had specifically called for quantifiability as opposed to mere spread of discussion. I think we should consider both.

• "This polling began as an informal effort by PMAnderson to see where we agree and where we disagree. It apparently morphed into a vote on what the MOS should say about dashes, but without any formal announcement of such."

I agree. Ill-planned, or not planned, from the start. But voting (with comments) grew naturally out of two weeks of discussion of what people wanted to talk about over the subsequent six weeks. Anyone could have attempted to re-focus discussion. Myself, I take it that the role of MOS as Wikipedia reference-point for punctuation was not brought into question or discussion. I therefore take this as accepted: for titles as for text.

• "What’s rational and consensual is sometimes in the eye of the beholder. If I were to summarize the archived discussion linked above, I would say that 1), there clearly was no consensus, and 2), the support for the current wording was primarily from two editors who pushed their preference pretty hard."

That's the real world, Wikipedia-style. Everyone had every opportunity to raise awareness and to steer discussion. I did all I could to extract a workable version of the guidelines we discussed, for real use in real editing. Nearly everyone approves – well enough! I have said to you several times that I myself disagree strongly with some provisions in the draft. Everyone does! That is not the point. Users of CMOS have their own private sets of gripes with it; so do users of NHR. That's how it is, with manuals of style. There can be consensus to adopt it, even if no two editors agree about which details are good and which are anathema.

• "I don’t suggest that the proposed practice is irrational, but simply that it’s no more rational than North American practice—and the latter avoids several problems, so in some cases, the current and proposed practice is irrational."

Many editors find it workable; I am an experienced editor, and I am among them. And don't speak of "North American practice" as if there were one. There are several. Yours and Dick's, for example, appear to differ considerably.

• "I would finish my comment here by saying that I see closing up range/etc. dashes as affording no disadvantages whatsoever in comparison with spaced usage in some cases, and it does afford a few advantages—closed-up usage largely precludes confusion with parenthetical en dashes and minus signs, and it avoids ambiguity in the instances of two dates followed by the year of the ending date. So at worst, it does no harm, which is essentially what I said to PMAnderson and a few others in cases where the en dash was seen as having little value. To me, that makes North American practice clearly more rational."

And I find otherwise, and some North Americans find otherwise regarding the various practices in your continent. We've had weeks to sort all this out. I've done my best to compromise, and to follow all that was happening. Sorry if you are not satisfied; let's all live with the ways we are not satisfied.

• "I haven’t bothered with formal draft of these points because I think I’ve made it quite clear what I would propose, and without reason to believe that such a draft would get serious consideration, I don’t wish to spend the time or effort."

I said explicitly (twice) that I needed to see a draft. I said that it would deserve consideration. You "don’t wish to spend the time or effort" of drafting a small section, where I have laboured to produce an entire draft, and revise it at least twice, and discuss it, and fend off calumnies without end, and endure prolonged last-minute reiterations of issues? Hmm.

• "I hate to be the one in opposition, but I simply cannot join an 'across the board consensus' that I cannot see."

You go on to say that we could come up with a partial consensus, for now. That was not our brief. We must work with realities, and with the task we face. I'm sorry you were not able to do that; but it is completely beyond my control – which is OK!

Jeff, thanks for your efforts. I will press on and post a draft that will bring this chapter of a monumental and dreary epic to a close. The big picture concerns the whole Project, and its quality. Let's all continue to focus on that, in the ways they we each see fit.

Noetica^Tea? 10:29, 23 July 2011 (UTC)

I strongly disagree on most of these points. I doubt that going through them would accomplish anything, so I shall not do so. JeffConrad (talk) 04:57, 24 July 2011 (UTC)

Answering Colonies Chris and OhConfucius

Thank you! Much appreciated. Expressions of support, suggesting that we just go ahead and finish it, with all that hard work behind us, are enormously helpful right now. I hope there will be some more like that.

Noetica^Tea? 13:00, 21 July 2011 (UTC)

Endorsements, and final points to fix (II)

[Editors, please see instructions for the first part of this "endorsement" phase, above. And continue here. I will respond further in 24 hours, adapting to small last minute suggestions. Noetica^Tea? 13:00, 21 July 2011 (UTC)]

• Ship it—Thanks Noetica! ErikHaugen (talk | contribs) 17:36, 22 July 2011 (UTC)
• Ship it No manual of style in a cross-country venue where *anyone can edit* can make every editor happy with every detail. Clearly, Noetica has done an exemplary job at trying to meld something that around which a consensus can form. Greg L (talk) 22:25, 22 July 2011 (UTC)
• Discard it and replace by Kotniski's draft, which is more accurate and less wordy; the proper word for this draft is "execrable". All comments on points of detail, including Kwami's and JeffConrad's, have been blown off; half of us (in two separate polls) have opposed these requirements; this doesn't reflect the usage of reliable sources. ---- (This post was by PMA on 01:05, 23 July 2011, note left here by Greg L; delete if/when this is signed,)

"Sidebar" issues, arising from recent discussion above

[Refactored. Editors, if you address substantial matters that are incidental to the main business, please do so in the style I use below. It will be very hard for people to find their way otherwise. Noetica^Tea? 00:01, 22 July 2011 (UTC)]

Sidebar on hyphen-related stuff

Somewhat off-topic, but shouldn't we have a spacing comment for slashes too? (Not here, of course.) I was reminded of this with the new Plutonian moon, which has been listed with both of its provisional designations as "P4 / S/2011 P1". The spaces around the first slash are obviously necessary, and AFAIK generally the convention whenever one of the elements being joined by the slash itself contains a space. — kwami (talk) 18:36, 20 July 2011 (UTC) Continued inside the navbox. Click on "show" at the right:

Sidebar on format of "correct" and "incorrect" examples

May I raise 2 related points about the format this draft section uses to distinguish between "correct" and "incorrect" examples?   (1) The style appears to depart significantly from that employed in many other sections of the MoS; and (2) the draft's (new?) formatting convention (color, font, and asterisk vs. no-asterisk) isn't explained anywhere (at least that I could find readily). ... [Posted by Jackftwist, and continued in the navbox below.–Noetica] Click on "show" at the right:

Jack, I believe that at least one participant in the present discussion cannot distinguish red and green. Yes, the important issue you raise came up earlier; the asterisk is simply to get us through for now. The matter needs to be revisited – later, may I suggest? Your input will be valuable! Incidentally, are you in a position to add your endorsement explicitly above (it is hinted at in your post), for proceeding with the present draft? The clearer we are about this the better. Noetica^Tea? 00:01, 22 July 2011 (UTC)

My own belief is that the MoS should be phrased in the imperative. We are telling people what to do, not describing what is done elsewhere. This style guide has a luxury that most do not: the ability to link to separate articles that can describe the prevalence and history of each rule, mark and space in detail. Darkfrog24 (talk) 00:53, 22 July 2011 (UTC)

Noetica: (1) I hereby bestow my nihil obstat and imprimatur for proceeding with the present draft, conditional only upon the one proviso and caveat noted below. By all means, proceed "with all deliberate speed." [The rest of my post is in the collapsed section.] — Jack --Jackftwist (talk) 18:50, 22 July 2011 (UTC)

Satisfying a {{who}} tag

Assuming consensus approved, good faith language, can the provision of WP:V citations in response to the placement of a {{who}} (or similar) tag be considered a WP adequate resolution to a WP:WEASEL objection? Thanks. JakeInJoisey (talk) 13:08, 25 July 2011 (UTC)

Not by itself in all cases. The language would need to support the source. So "some say ...." would need to be replaced with "in an editorial in The Times said....." if that was your only source. Something like "It was widely said at the time...." could stay, if it was sourced to a source that analysed the coverage and said "it was widely said at the time....". If the answer to "who" is one person/source only, then (1) the wording should be altered to show this and (2) WP:UNDUE may come into play. If the language itself can be reliably sourced, then it may not be necessary to alter the language. If it is an attempt to evidence "most people" by citing half a dozen people who said it, then that may be WP:OR - and WP:NPOV if there are another half dozen sources that contradict it. --Elen of the Roads (talk) 13:39, 25 July 2011 (UTC)

Thanks for your consideration and I am in full agreement with your position though I'm not clear that you may have adequately noted my quite purposeful use of the plural "citations".

To be specific and if I interpret your opinion correctly, in cases where multiple, diverse RS sources more than satisfy consideration under WP:V and consensus agreement exists for the language employed, source citations CAN, in your opinion, satisfy a {{who}} objection, correct? JakeInJoisey (talk) 13:58, 25 July 2011 (UTC)

Sorry, did not catch that you were being specific in citations. In general, good sources can certainly satisfy a who objection. If the sources support the language, WP:WEASEL should not be an issue, as the main concern of weasel is unsourced non-specifics - vague wording intended to camouflage poor research, or the use of words that have a different meaning than the source supports.--Elen of the Roads (talk) 16:01, 25 July 2011 (UTC)

That is succinct and precisely the clarification I was seeking. Many thanks for that.

If I might note an ancillary question I've posed, you might find Template talk to be of some interest as well. Thanks for your consideration. JakeInJoisey (talk) 16:45, 25 July 2011 (UTC)

In addition, if one has a source, one might as well say "According to Brutal Copy Editing, people who use apostrophes to create plurals are 'morons.'" Situations in which the structure of the sentence or paragraph would prevent the addition of such information would be rare. Darkfrog24 (talk) 17:55, 25 July 2011 (UTC)

...if one has a source...

A single source is, I'd suggest, an entirely different consideration than an availability of multiple sources...as discussed above, but I certainly concur that a single source {{who}} objection is easily addressed by attribution. JakeInJoisey (talk) 18:11, 25 July 2011 (UTC)

The solution to "too many cooks" is to make one of them head chef: "According to several high-end style guides, including Brutal Copy Editing, people should not use apostrophes to make plurals." It would be like "et al." in a scientific journal. Darkfrog24 (talk) 21:47, 25 July 2011 (UTC)

Short answer: Yes. A weasel tag does not require WP:INTEXT attribution of a fact or opinion to any named source.

Related to Darkfrog's example, if a substantial proportion of (relevant) sources support the statement, and none contradict it, then it is actually inappropriate and misleading to call out any particular one as holding the opinion. We assert widely agreed upon facts without specifically naming any particular person who holds that opinion. "According to many mathematicians, including Alice Expert, two plus two equals four" would be silly.

Or, to give the more relevant example, if dozens of media sources say that the Casey Anthony trial is a media circus on par with the OJ Simpson trial—which they do—and none of them contradict it—which they don't—then we might well provide a string of sources, but we're not going to single out any one of them as especially holding this essentially universal view. WhatamIdoing (talk) 23:03, 29 July 2011 (UTC)

Interesting article on the Associated Press (AP) Stylebook

You might be interested in (and possibly amused by) this article from the "Style" section of the Friday, July 29, 2011, edition of The Washington Post** about the AP Stylebook and its chief editor.

http://www.washingtonpost.com/lifestyle/style/david-minthorn-is-the-grammar-expert-for-the-associated-press/2011/07/25/gIQAGBLwfI_story.html

Caveat: washingtonpost.com is notoriously s-l-o-w to download, so click on the link and find something else to do for a few minutes (potty break, coffee refill, etc.)

If you take the stylebook knowledge quiz, note that it's testing only knowledge of the AP's guidelines, so some of its "correct" answers may be "incorrect" according to other guides. (I.e., even the most experienced WP editor shouldn't necessarily expect to score 100% on it.)

The print edition's headline for the article was a clever graphic, with the following text (in huge font), complete with copyediting marks correcting the typos and deviations from AP style (ahh, the occasional advantages of print over HTML):

Make no mistake; The folks behind the "AP style book" are

choosey about words such as 'e-mail,' 'octopi' and 'smart phone' *

Corrections shown by the copy editing marks:

• colon instead of semicolon after "mistake"
• "Stylebook" is 1 word, capitalized, no quote marks around the title and not italicized (that's the AP rule for "standard reference works" and the Bible; they use quotes around other titles and never use any italics)
• "choosy"
• The * refers to the following footnote about AP's guide: "They prefer 'email,' 'octopuses' and 'smartphone.'"

** Note: The Washington Post 's stylebook requires that "the" be capitalized and included as part of its name, which seems to be a common (but not necessarily universal) practice among newspapers whose names begin with "the." All punctuation in the headline follows standard U.S. rules (e.g., commas inside quote marks). Apparently, both The Post's and AP's stylebooks omit the serial (terminal) comma in series of 3 or more items. Enjoy. --Jackftwist (talk) 21:34, 29 July 2011 (UTC)

RfC for Collapse boxes on chess problems

Recently on Susan Polgar, the issue was raised as to whether chess problems (and, presumably, other problems where readers may prefer not to read the solution until they've tried to figure it out themselves) constitute a reasonable exception to WP:COLLAPSE. The issue was discussed a bit on Talk:Susan Polgar and then it moved to Wikipedia:Administrators' noticeboard/Incidents#Susan Polgar Hiding in plain sight. It would be helpful to see the input of uninvolved editors, and, if consensus is that this is an acceptable use of hidden text, consider adding in that exception either here and/or at WP:Spoiler. Qwyrxian (talk) 04:38, 23 July 2011 (UTC)

Why should we make this exception to WP:SPOILER? Though, on the other hand, it's not exactly the same as a spoiler warning, or not including spoilers. The content would still be there, just kind of hidden. And it's a reasonable courtesy to people who want to learn about chess but not have the tactics ruined for them. So I don't see why not. hare j 05:14, 23 July 2011 (UTC)

No guideline is intended to apply to all possible situations. I often object to those who invoke WP:Ignore all rules to ignore a guideline altogether, but not when they only assert an unwritten exception. Requiring a chess problem to show its solution, makes it almost useless. If the guideline was followed but the content was useless, that's like saying the operation was a success but the patient died.

It should also be mentioned that there was a lot of unnecessary hostility. But you've already been to Wikipedia:Administrators' noticeboard/Incidents to handle that problem. Art LaPella (talk) 05:26, 23 July 2011 (UTC)

Technically I am an uninvolved editor to the extent that I joined the discussion at AN/I. If that in turn makes me an involved editor, then consider the following comments as being from someone who has played chess since the age of, oh, maybe 5. (I wasn't that good, I am certain my aunt let me win on purpose.)

In giving this opinion, I will link to some chess puzzle sites. This is merely a list; I will go into detail under this list.

• A simple chess checkmate puzzles site (intermediate puzzle #1)
• ShredderChess
• chessproblems.com

Keep in mind, this is based on a quick Google search.

Looking at the first site, it's your basic here's the picture, scroll down for the solution. This is an old school way of doing things. In a book, you have absolutely no choice. The Internet is not a book, and we should not treat it that way. Moving to the second, it's daily puzzle (the easy one for today is LAME, BTW - no instructions and I was trying to checkmate only to be told the goal was to take the Bishop) appears and you can click and move the pieces. It's a basic interface, click and click again. The third one is beautiful; you actually click and drag the piece, and the system plays moves in response. If you give up, there is an option on the left to animate the solution. (I had to do that for the one I tried).

Now, two of these do not have the solution on the page at all. This is the ideal way to do it; keep the solution to such a puzzle hidden from the vic-...*ahem* participant and only reveal it when asked. The solution with which I agree to at Susan Polgar when presented with Polgar's first chess problem (composed at age 4) was to have the solution hidden, but when the article reader looks at the diagram and makes a decision on the solution, they can then click the link to see if they're right. They can't play on the board directly (we are not a chess site, why would we?), but they can certainly try the puzzle without the chance of seeing a solution immediately by accidental scrolling (as could be the case in site #1, and some screen sizes might let users see the solution immediately without scrolling).

The question was raised: why not put it in a separate section? No good at all; the last thing we need in an encyclopedia article is a section titled Solution to above problem. It's ridiculous in an online setup. In a paper encyclopedia, sure, it's great, but in this format it is just a mess and not something that needs its own section. Hiding the solution right under the grid and allowing the user to click when they are ready to see the solution is an ideal sol-...er, is the ideal way to handle this kind of presentation. CycloneGU (talk) 05:35, 23 July 2011 (UTC)

What he said. Swarm 06:15, 23 July 2011 (UTC)

I definitely think this is a reasonable exception to the WP:COLLAPSE guideline, which is, after all, only a guideline, and I don't think WP:SPOILER is 100% applicable to this situation. In the ANI discussion, it was stated that concealment of the solution is, in a way, part of the content. I think that makes a lot of sense. It could be thought of as analogous to an optical illusion. The illusion is the content. —Torchiest ^talk_edits 06:37, 23 July 2011 (UTC)

While I don't think that what other sites do is relevant (because we have a different mission than a chess site), Torchiest's argument that the hidden-ness is essentially part of the content itself does make some sense to me. I'm still not fully convinced, but at least I can see a context for justifying this but not other types of spoilers. That, by the way, really is my biggest concern--I don't like the idea that others can use this as precedent to say "that local group argued there situation is special, but ours is also special so we should be able to hide our information as well". This argument sets up the idea that the difference lies in the actual material itself; it's almost like, if we were "quoting" these problems, we would have to separate the solution from the question, and our technical way of doing that is with hidden text. I'd still like to hear wider input, especially from MOS-gnomes, who can approach this not from a concept of what's appropriate for Chess problems, but how this fits into the greater logic of the MOS. Qwyrxian (talk)

The point of my demonstration was to show different styles of laying out this type of problem, not to advocate a certain site's layout over another. After all, we are our own site, we do our own thing. I merely use the examples given to demonstrate why having the solution hidden (which itself is a spoiler to the puzzle) is the way to go, and two examples of sites I gave let you click a button to see the result if you give up. That's a similar format to what the AN/I agreed on here with a collapse box. CycloneGU (talk) 14:37, 23 July 2011 (UTC)

• I think the issue can be avoided: at the article talk, I have suggested that a standard numbered footnote be used to provide the solution. That is much more in keeping with standard procedures. Johnuniq (talk) 09:38, 23 July 2011 (UTC)

  That's only satisfactory if footnotes exclusively containing bibliographic info are kept separate from those containing content, as in quark#Notes. Otherwise, it might not occur to all readers that the footnote contains the solution rather than a citation. ― A. di M.​_plé​^dréachtaí 19:06, 23 July 2011 (UTC)

• My view is that the collapse box is acceptable for solutions to any kind of incidental puzzle. Were the article about this problem (and I am no chess expert but it seems a pretty trivial problem to me) then a collapse box would be inappropriate. In this article I think it works - and it doesn't just have to apply to chess, either. Although I admit I'm struggling to think of other non gaming examples where it would apply! Egg Centric 12:43, 23 July 2011 (UTC)

  Egg, yes, as far as chess problems go, it's pretty trivial. But!, were you aware that Susan was 4 years old when she composed it? Ihardlythinkso (talk) 14:48, 24 July 2011 (UTC)

• There is a fundemental difference between a collapse box in an article that describes a book and one that describes a gaming puzzle - in the former, the reader is unable to ascertain the answer without reading the book (or opeing the box) whereas in the latter, the reader can find the solution without having the relavent gaming board present. This suggests to me that the guidance should be modified to council against collapse boxes in spoiler situations only in circumstances where the reader would be expected to consult an outside work to find the answer. Martinvl (talk) 13:03, 23 July 2011 (UTC)
• What implications does this sort of thing have for accessibility issues, i.e. for the visually impaired? LadyofShalott 15:52, 23 July 2011 (UTC)
• Don't collapse puzzles; maybe put solution at article bottom - The WP:collapse guideline is very clear. Readability is paramount, and WP is supposed to contain prose-style encyclopedia articles. Many readers may not understand what "show" or "hide' means. Especially non-English speakers, or people with accessibility issues. Presenting information is far more important than avoiding spoilers. Plots of movies and books are clearly presented in WP articles, without any "show/hide" mechanism. See WP:Spoilers. See also Wikipedia:No disclaimers in articles which explains why spoiler alerts are not included in any WP article. As a compromise, consider the approach used in Chess problem, where there is a puzzle presented in the center of the article, and its solution is located at the bottom of the article, and a link jumps the reader from the puzzle to the solution. --Noleander (talk) 18:21, 23 July 2011 (UTC)

  Disagree with view - If I am reading an article, I want the solution to be near the puzzle (not just chess but any theoretical puzzle used properly within the article), but not necessarily visible so as not to ruin the concept of the puzzle. The collapse option makes perfect sense here and is a suitable exception. This RfC is suggesting to make changes to the guideline (and note it's a guideline, not a policy, and cannot be enforced as such), and just responding saying that "the guideline is very clear" is completely ignoring the fact that this is a request to introduce changes to the guideline. CycloneGU (talk) 20:18, 23 July 2011 (UTC)

Following up on Lady of Shalott's question: I took a look at Wikipedia:Manual of Style (accessibility)#Users with limited CSS/JavaScript support. At first this worried me, so I tried to look at the page with Javascript turned off. Conveniently, the page displays all of the information (basically, the way the article looked before the hidden line was added), so at least we don't have to worry about that. Regarding the visually impaired...I strongly believe that we should do what we can to make their browsing of Wikipedia as functional as possible...but in this case, isn't it correct that no matter what we do, the visually impaired can't use the problem any way, since it's contained in an image? In which case, it makes no difference how the solution is portrayed. Qwyrxian (talk) 00:51, 24 July 2011 (UTC)

That can be handled by means of alt attributes. Right now the template gives alt texts like “white queen” etc., but it might be tweaked to read “white queen in h4”. ― A. di M.​_plé​^dréachtaí 10:47, 24 July 2011 (UTC)

Just on that point, it is not impossible to give information to a person with a visual impairment without using the image. It just takes more effort. We should try to do that if at all possible. Regarding the collapse box, I think a screen reader 'reads through' the box, similar to the effect with Javascript off. Elen of the Roads (talk) 16:57, 24 July 2011 (UTC)

OK, thanks, y'all, for the accessibility input. I do think the alt text should be made as informative as possible a la A. di M.'s suggested tweaking so that people using screen readers can obtain full benefit of the article's content. LadyofShalott 20:19, 24 July 2011 (UTC)

• As a chess player with a passing interest in chess problems, I went and had a look, and while I appreciated the solution being hidden here, it was indeed a rather trivial problem to solve (the interesting thing is working out why a White piece is needed on d2 and why it has to be a pawn - the answer being that without that pawn there, there are two solutions to the puzzle, which would be a 'cook' - a failed problem - and that other White pieces on d2 would allow other mates). Personally, I think the solution should be in a footnote, but I'm not sure how chess problems and endgames (or indeed positions from notable games) should be handled. I'd be inclined to just present them without solutions! :-) Carcharoth (talk) 21:51, 24 July 2011 (UTC)

Personally, since there don't seem to be any accessibility issues, I prefer the hidden text to the footnotes. This is because I really don't like mixing "informative" footnotes with "citation" footnotes; to me that's extremely confusing, as when I see a reference marker, I assume that it leads to a citation, not to a place with further info about the text. Yes, there is a way to help avoid this confusion, by using reference groups, but I think that that is quite difficult for new editors to learn how to use, and still isn't entirely clear to a reader. If hidden text is the preferred option, what would people think of changing

Collapsible sections may be used in navboxes or infoboxes, or in tables which consolidate information covered in the prose.

to instead read

Collapsible sections may be used in navboxes, infoboxes, tables which consolidate information covered in the prose, and solutions to puzzles or problems (such as chess problems) whose solutions are normally hidden or distanced from the problem in reliable sources that collect such puzzles.

That's probably too wordy; my intent was to distinguish puzzles which hide the answer as solving it is part of the "fun", to mathematical texts, which don't hide the answer to their "problems" except in textbooks, where the purpose is different and hiding is definitely not compatible with the way Wikipedia works. Also, I would like it to be very clear that the exception only applies to this particular class of problems, not to plot spoilers, to offensive material, etc. Qwyrxian (talk) 00:20, 25 July 2011 (UTC)

I like that. How about:

Collapsible sections may be used in navboxes, infoboxes, tables which consolidate information covered in the prose, and solutions to puzzles or problems (such as chess problems) normally hidden or distanced from them in sources that collect such puzzles.

Slightly trimmed down, although maybe someone else could do even better? —Torchiest ^talk_edits 13:32, 25 July 2011 (UTC)

My go:

Collapsible sections are permitted in navboxes, infoboxes, tables which consolidate information already covered in the prose, and solutions to puzzles or problems (such as chess problems) normally hidden or distanced from them in sources that collect such puzzles.

Notably, I kept the s as part of the Wikilink on problems and chose a word I like better for a guideline. CycloneGU (talk) 14:38, 25 July 2011 (UTC)

Doesn't putting the 's' outside of the link but right next to it automatically make it part of the wikilink anyway, without needing the piping? Either way, though, the effect is the same. I like the simplification of the final clause. I have no opinion on "may" vs. "permit", which seems to be the subject of a significant discussion below. Qwyrxian (talk) 01:45, 27 July 2011 (UTC)

Based on an above example, it appears that is indeed the case. I wasn't aware the code recognized that. CycloneGU (talk) 13:24, 27 July 2011 (UTC)

• As an uninvolved editor, I strongly support the idea of making the suggested changes to WP:SPOILER and/or the Manual of Style. Chess puzzle solutions should be hidden in some way - that's exactly what I would expect if I were reading a book about chess, or even a bio about a notable chess master. Wikipedia needs to be flexible enough to allow various WikiProjects to make exceptions to guidelines and even policies, if need be and when appropriate. A collapsible section seems to be the most elegant way to do this. First Light (talk) 04:41, 29 July 2011 (UTC)
• Personally, I think that SPOILER has it right. All collapsible boxes unavoidably cause one accessibility problem: Any person who finds it painful to click on things (e.g., due to carpel tunnel syndrome) has to choose between experiencing pain and not reading the full article. WhatamIdoing (talk) 22:40, 29 July 2011 (UTC)

  I don't see anything in WP:SPOILER about carpal tunnel syndrome or other disabilities. And I do see hide/show boxes everywhere on Wikipedia. They are frequently on talk pages, and the first featured article I just checked for an example had sixteen hide/show navboxes at the bottom, with 'hide' being the default. Maybe the developers on Wikipedia could have an option for people with disabilities to have all hidden boxes default to "show". First Light (talk) 14:19, 30 July 2011 (UTC)

• Hide chess solution. I went to the Polgar for the first time and I was delighted that her four-year-old youngest-ever-published problem was open to my attempt to solve it without the answer being right there in front of me. All chess problem solutions should be hidden. Binksternet (talk) 14:54, 30 July 2011 (UTC)

• I'd like to point out, though of course I favor use of "hide"/"show" for chess problem compositions, and appreciate the language recently constructed for updating MOS, there exists a number of uses of "hide"/"show" already in some rather mature articles, for use not to depict chess problem composition solutions, but rather brilliant or difficult game continuations. For examples of these, please see articles Mikhail Botvinnik (contains two such game continuations), Wilhelm Steinitz (one), and Alexander Alekhine (one game continuation in "Notable chess games" section, plus a problem composition in the same article). And I'm sure there are other article examples besides these. In addition, besides game continuations and problem compositions, there is an additional use of "hide"/"show" demonstrated in the mature article Emanuel Lasker – a diagram about the Rice Gambit. My concern is, if the language currently under consideration to update MOS is drafted, will it be so restrictive for chess as to allow use of "hide"/"show" for only chess problem compositions? And put the kibosh on use for brilliant or difficult game continuations? Ihardlythinkso (talk) 20:34, 30 July 2011 (UTC)

  I think if it said "chess problems and continuations" it would be seen as covering all of those, unless you think more explicit language is needed (keeping in mind that I'm completely ignorant about chess, but wanting to see the language be both simple and comprehensive). First Light (talk) 03:46, 31 July 2011 (UTC)

  I think the phrase you're suggesting is excellent. But will others here agree to the less restrictive language? Also, another part of the language under consideration says "... normally hidden or distanced from them in sources that collect such puzzles." But many times that does not describe or apply to brilliant or difficult game continuations found in books (which often, if not usually, print the continuations on the same page, i.e. not "hidden"). So am wondering if further modification to the language will be accepted in this discussion, to not put the kibosh on brilliant or difficult game continuations already found in mature articles as exampled above.) Ok, Ihardlythinkso (talk) 05:55, 31 July 2011 (UTC)

  p.s. There are lots of books too, which do segregate position diagrams, from their brilliant or difficult game continuations (such as in the appendix). But I doubt one could call that the "normal" way it's done (the current language says "normally hidden"). There is the distinction, that brilliant or difficult game continuations come out of praxis/real games, whereas problem compositions are invented. But unlike problem compositions, brilliant or difficult game continuations sometimes are segregated from their diagrams, sometimes not. (The editors who included game continuations in Mikhail Botvinnik, Wilhelm Steinitz, and Alexander Alekhine made their choice to segregate or not, choosing to hide. To me that seems best – to leave to consensus what's best for an article. But would that degree of flexibility be accepted here when drafting language update?) To summarize my thought here: solutions are normally segregated from their respective problem compositions. Brilliant or difficult game continuations sometimes are segregated from their diagram positions, and sometimes not. If I had to guess, I'd say the assertion they are normally segregated in books, is false. (So if we accomodate game continuations in the language, the "normally hidden or distanced" part can't apply to them.) Ok, Ihardlythinkso (talk) 06:38, 31 July 2011 (UTC)

I'm fine with adding "continuations"; I'm not fine with the idea of Wikipedia articles being substantially different from our reliable sources. If they're not normally segregated there, I don't think we should be making an exception here. As Ihardlythinkso knows, I strongly dislike the idea that editors of individual articles just get to declare that a particular guideline doesn't apply to them. Otherwise, people could say, "I don't care if that author is from the UK, people in the US read her a lot so we're going to use American English despite WP:ENGVAR", etc. There's no point in having guidelines if they can essentially be overruled in every case by one or two "local" editors. Qwyrxian (talk) 12:31, 31 July 2011 (UTC)

Qwyrxian, I know you have worries and concerns, and that's normal for someone (now Admin) in your position, and the position you've assumed in this discussion, helping draft update language to the guideline. But I'm lowly editor, so I don't think or worry along the same lines as you, regarding uniform enforecment of WP policy, what precedents might mean re future implications, etc. So I can't help in that regard. About the issue here though, when you say you're "not fine with the idea of Wikipedia articles being substantially different from our reliable sources", as far as the game continuations I've mentioned (two in Mikhail Botvinnik, one in Wilhelm Steinitz, one in Alexander Alekhine, something on Rice Gambit in Emanuel Lasker), I'm not in position to answer that and say "yes they are" or "no they're not", because I don't know. And I don't know if those specific game continuations AND their respective diagrams are, or are not, in reliable sources. (Because what we're talking about here is really a pair – a problem composition + its solution; a real game position-diagram + its continuation.) I have no idea. Maybe the editors who added that material would know. But they aren't here discussing. There's no way to know without their participation. (Which just makes me more curious about this whole process. What language can or will be drafted? And what are the implications for the aforementioned articles containing game continuations? (And I'm sure there are many more like them!) I mentioned before how brilliant or difficult game continuations might be "hidden" (appendixed) in books sometimes, or probably usually they aren't and occur on same page. But for any particular game continuation in any particular article (like the aforementioned ones), what will be done? Independent investigations for each? Revert for each unless editors can prove those particular game continuations are found in a reliable source, equally hidden? I don't even know if all that research is possible or likely! It will certainly be time consuming. But if you like to be uniform, as you do, there's no ignoring all of this intricacy and difficulty. (And to think it all started with the cute, little Pogar 2-mover, when you discovered me applying "hide"/"show" to the solution!) I'm curious what will become of this RfC, in light of the intricacy and difficulty alluded to. And another amazing point: I'm not aware anyone from WikiProj Chess registering comment here, or in the AN/I, or even at Talk:Susan Polgar! (I've been also curious about that too. I wonder why it is. I wonder if they don't see any of this worth their time. And if they don't, why they don't.) When you reverted me at Susan Polgar, it was just because the edit caught your eye. While there are probably an unknown, maybe hundreds, of chess problem compositions already existing in mature articles. And now that I've introduced the point of brilliant and difficult game continuations, also found in mature articles, it just expands the issue very broadly, mutiplying the intricacy and complexity and difficulty. (Does that explain the lack of participation from WikiProj Chess members? That they view it, if they took a look at all, as nothing more relevant than Ihardlythinkso being picked on at Susan Polgar? A funny insignificant trifle, since the broader spectrum of already-imbedded problem compositions + solutions, and real game positions + brilliant or difficult continuations, are widespread and deep, and no one or two Admins who jumped on Ihardlythinkso could possibly pose a challenge to those mature articles, or would ever think of making new policy and then use it as a precedent to go messing with those articles, ever? So why give a hoot? [It's my latest theory. But I'm just thinking out loud here. Please ignore.]) Peace! Ihardlythinkso (talk) 13:50, 31 July 2011 (UTC)

I would still like to see some more participation from WikiProject Chess, but barring that I think that making "chess problems and continuations" an exception is an elegant solution. I also see no harm in doing so - the "hide" default is used so widely for other purposes, this just shouldn't be that big of a deal. First Light (talk) 15:02, 31 July 2011 (UTC)

First Light, thx, but I doubt Qwyrxian would agree w/ you. Ok, Ihardlythinkso (talk) 17:10, 1 August 2011 (UTC)

En dash change proposal

If you're weary of en dashes, and don't care exactly where to draw the line on the compromise about when to space them in complicated ranges, read no further. If you care, let's consider adjusting it. Jeff and I had a disagreement about process, but I pretty much agree with some of his points, and would suggest pushing Noetica's artful compromise just a little bit further. Jeff seems to have declined to provide a specific suggestion, so I'll propose one.

The present version says:

---

The en dash in a range is always unspaced, except for times and dates (or similar cases) when the components already include at least one space.

• *23 July 1790–1 December 1791;    *14 May–2 August 2011
• 23 July 1790 – 1 December 1791;    14 May – 2 August 2011
• 10:30 pm Tuesday – 1:25 am Wednesday;   Christmas Day – New Year's Eve;   Christmas 2001 – Easter 2002
• 1–17 September;   February–October 2009;   1492? – 7 April 1556
• Best absorbed were wavelengths in the range 28 mm – 17 m.

---

I propose to take out some of the positive examples, and reword the guidance to narrow the situations in which spacing is used:

---

The en dash in a range is always unspaced, except when the ends of the range are long enough or complicated enough, with spaces in them, that it is important to signal a wider scope of connection; for example, to avoid connecting a number with one component directly with an inappropriate part of the other.

• *23 July 1790–1 December 1791;    *14 May–2 August 2011 (inappropriately suggesting 1790–1 and May–2)
• 23 July 1790 – 1 December 1791;    14 May – 2 August 2011
• 10:30 pm Tuesday – 1:25 am Wednesday;   Christmas 2001 – Easter 2002;   1492? – 7 April 1556 (avoid suggesting Tuesday–1:35, 2001–Easter, and 1492?–7)

Having a space, or mixed numbers and words, in a component is not necessarily enough to trigger this need; editorial judgement is needed, as simpler components with spaces in them should retain the en dash unspaced:

• wavelengths in the range 28 mm–17 m;   Christmas Day–New Year's Eve (mm–17 and Day–New are unlikely to be confusing)

Sometimes recasting to avoid the complication is preferable:

• wavelengths in the range 0.028–17 m;   from Christmas Day through New Year's Eve

---

Please support or opppose this change, and/or propose a modification to make it better.

• Support as nom. I don't care strongly one way or the other, but if Jeff is right that there is more support, or more reasoned support, to less often spacing the en dash, maybe that will come out here. Basically, I think that choosing among the options suggested in style manuals is really an "I like it" process, in spite of what he says, so I have nothing more to say except that I think I'll like the result of this guidance a little better. I won't be offended if others prefer to keep what Noetica came up with. Let's don't make it a fight, just see which way the wind is blowing. Dicklyon (talk) 21:43, 24 July 2011 (UTC)

• Comment I tried to stay out of the details while we were trying to actually once again function like civilized hairless apes. But some of these complex ranges look like poor writing practice to me. I think there is too much symbolitis on Wikipedia. Shall we consider something like the following:

---

For simple ranges, the en dash is unspaced.

• The usual amount added to Olympic-size swimming pools is 20–30 liters.

Spaces on both sides of the en dash are often best for mixed measures and those with multiple components to the measures:

• *23 July 1790–1 December 1791;    *14 May–2 August 2011 (inappropriately suggesting 1790–1 and May–2)
• 23 July 1790 – 1 December 1791;    14 May – 2 August 2011
• 10:30 pm Tuesday – 1:25 am Wednesday;   Christmas 2001 – Easter 2002;   1492? – 7 April 1556 (avoid suggesting Tuesday–1:35, 2001–Easter, and 1492?–7)

Editors should consider recasting complex measures; particularly where space is not at a premium (i.e. body text and not tables). Mixed measures or those with multiple components to the measures are often more readable fully written out:

• wavelengths in the range 28 mm–17 m;   Christmas Day–New Year's Eve
• wavelengths ranging from 28 millimeters to 17 meters;   from Christmas Day through New Year's Eve
• from 23 July 1790 to 1 December 1791

---

That’s my suggestion. Greg L (talk) 22:22, 24 July 2011 (UTC)

Thanks; I took out a parenthetical that you left from mine that was inappropriate in yours. Let's hope future commenters will tell us which version they prefer (or OK with either); I'm OK with either. Dicklyon (talk) 22:36, 24 July 2011 (UTC)

• Support. (merged after many ECs) A few suggestions:
  1. I’d like to see some way of dealing with 24 July – 3 September 2012, for which I’ve previously noted that the beginning year is ambiguous. If it is meant that the start year is 2012, I think 24 July–3 September 2012 would handle it just fine, though I’m sure some others would disagree. I see no problem whatsoever, and at worst see it as a lesser evil. I recognize that this would raise some other issues (e.g., if this were acceptable, why would not 24 July – 3 September also be?). At the risk of hammering a point, I might also mention that it finds support in arguably the majority of top-selling guides, so the practice is hardly off the wall; I’ll save the clutter of mentioning them unless someone asks. Of course, some other widely used guides sometimes suggest the spacing, so perhaps this could be an area of editorial judgment (read: elements of our common language that aren’t common). One could of course eliminate the ambiguity by always including the start year, but this seems unnecessarily burdensome.
  2. I’d like to mention considering recasting wavelengths in the range 28 mm–17 m to wavelengths in the range 28 mm to 17 m or wavelengths in the range from 28 mm to 17 m, just because so many technical organizations fear confusion of the dash and minus (and apparently some of us aren’t thrilled with the dash, spaced or unspaced). I’d also like to see examples of both nominal and adjectival usage (e.g., wavelengths in the range 28 mm to 17 m and wavelengths in the 28 mm–17 m range); with the latter, recasting to use to would be difficult. The same might hold in a table where space is at a premium.
  3. The guiding principle should be to minimize ambiguity, though the approach ultimately chosen may vary with editor and situation—it’s simply impossible to cover or even anticipate every situation. I certainly fell a bit short in the polling breakout.

JeffConrad (talk) 22:45, 24 July 2011 (UTC)

• I’m ahead of you on this one and agree. As I proposed above, these complex ranges are best just written out. Check it out. Greg L (talk) 22:57, 24 July 2011 (UTC)

Hey, I started mine before you posted . . . perhaps great minds think alike. JeffConrad (talk) 23:06, 24 July 2011 (UTC)

Seeing a quick general consensus on something that seemed like a no-brainer, I took a close version of what I proposed above and grafted into MOS here at this point (#1, where it begins with “For simple ranges with a single unit of measure”…). There is a slight redundancy now, but I think the point it makes (getting editors to think about whether they have two different units of measure to the same measure or have multi-component units) is worth providing. Greg L (talk) 23:26, 24 July 2011 (UTC)

You're more bold than I! I expect we'll get to discuss some more after you get reverted, but maybe I'm wrong and it will be OK. I once made an MoS change after brief discussion; it was a good way to find the opposers, and went down in flames. Dicklyon (talk) 04:55, 25 July 2011 (UTC)

I have problems with this bit: "Spaces on both sides of the en dash are often best where there are mixed units for the measures as well as those with multiple components to the unit." First, you don't give an example of "mixed units", and probably don't mean to suggest doing the spaces in "28 mm – 2 m" and such. Second, I don't what "multiple components to the unit" is supposed to mean. Can you fix? Dicklyon (talk) 05:02, 25 July 2011 (UTC)

Indeed, I was not clear. Measures are things like volume or time. Units of measure are things like "kilogram" and its symbol kg. After your post, I checked my booleans and text and realized what I had there was unnecessarily confusing. I revised it so it should be much improved. And it now uses parentheticals to clarify buzzwords like “units” and “values.” It is here in this version of the en dash section, beginning with For simple ranges where the en dash replaces the word “to”… If you don’t like what I have there, please revise until it makes sense to you. Greg L (talk) 15:14, 25 July 2011 (UTC)

Your revision for clarity seems to have flipped advice back to more spaced en dashes, in 28 mm – 17 m. I thought this was what we were moving away from. But I'll let you and Jeff work it out for a while, and see if anyone else cares. Dicklyon (talk) 03:12, 26 July 2011 (UTC)

As for Jeff's corner case "24 July–3 September 2012", I can't imagine how anyone could interpret it as other than two dates in the same year, whether the dash is spaced or not. Anyone who wants to say from about now to a date later in 2012 would add the 2011, if they had any sense. So I would oppose any attempt to "deal with" this case by way of punctuation. Whether you decide the two dates with one year should have spaces or not is no big deal to me. Dicklyon (talk) 04:12, 26 July 2011 (UTC)

While we're at it, why not a date example with commas in it: (July 24, 2011 – September 3, 2012)? I see these in bios and such. Dicklyon (talk) 04:17, 26 July 2011 (UTC)

On ranges such as “28 mm – 17 m”: same impression as Dick. Again, I would show both nominal and adjectival usage, because the latter is more difficult. In many cases, the adjectival usage can be recast to a nominal, but sometimes it cannot be done easily. My example of “20 kHz–20 Hz propagation delay” might be recast as “propagation delay from 20 kHz to 20 Hz”, but this also is initially confusing because it reads as if propagation delay is reduced from 20 kHz to 20 Hz. Clearly, this is nonsense and should be quickly recognized as such, but there is a tiny bit to sort out, so it makes for more difficult. Any way, the phrase is jargony, and possibly confusing to anyone who isn’t pretty well versed in what is being discussed. But “the propagation delay of a 20 Hz signal relative to that of a 20 Hz signal” is pretty awkward, too. In this case, I think the closed-up adjectival use is arguably the least of evils. And in a table, it might be the only option.

On “24 July–3 September 2012”: you’re probably right—it’s more a case of initial ambiguity, though spreading things out does tend to loosen the association of “4 July–3 September”. It’s not as if most of these things can’t be sorted out, but rather at how much of a double take is needed. Quite honestly, I have a tough time imagining anyone who sees “4 July–3 September” reading “July–3”; after all, the spacing is nominally half an em vs. a third em, and “July–3” would be a mighty unusual way of giving month + day. But I guess it depends on the main determinant of association—is it the amount of separation or the type of mark? There definitely appears to be two schools of thought on this.

In any event, ranges of physical quantities are a bigger deal because at least some of us frequently deal with addition and subtraction of them. What think you, Greg? JeffConrad (talk) 07:15, 26 July 2011 (UTC)

• Jeff, I was trying to edit within consensus. My read about consensus was that other editors here wanted to replace the “to” in date ranges and the trick was how best to make it most readable, thus 23 July 1790 – 1 December 1791.

  Personally, I think this and similar constructions like wavelengths in the range 28 mm – 17 m read very poorly. The whole point about all writing is to make it enjoyable to read. The great challenge of technical writing is to make it as non-confusing and illuminating as possible for the target readership. Any text or writing style that unnecessarily slows down the eye or unduly calls attention to itself is verboten.

  Many readers live in countries where dates are commonly formatted a particular way. For instance, the date example I used here uses a convention that is common in European countries. Americans, who are accustomed to July 23, 1790 already have their brains slowing down to parse formats such as this; adding an en dash makes the brain work even harder to parse the construction.

  The same brain struggle occurs when we mix different units of measure. I personally think that wavelengths in the range 28 mm – 17 m is an abomination unto the eyes of the technical writing gods. When I ran across this the very first time here, my eye doubled back twice and my first thought (or second or third—I can’t recall) was that a mistake had been made and an “m” had been omitted. Is there a respected manual of style that we pulled that one from?

  Were it me, my advise would be as follows:

---

For simple ranges where the en dash replaces the word “to” between two positive integers (numbers like 20 and 30), the en dash is unspaced.

• The usual amount added to Olympic-size swimming pools is 20–30 liters;   pp. 211–219;   64–75%;   the 1939–45 war

For date ranges involving years from 1000 AD or later, it is acceptable to truncate the first two digits from the second year. For all other numbers, do not truncate.

• Right: pp. 211–219;   64–75%;   the 1939–45 war;   the 1939–1945 war
• Wrong: pp. 211–19

Do not mix en dashes with prepositions like between and from.

• 450–500 people
• between 450 and 500 people, not between 450–500 people
• from 450 to 500 people, not from 450–500 people

If negative values are involved, an en dash might be confusing. Use words instead.

• −10 to 10, not −10–10
• the drop in temperature was from 5° to −13°, not the drop in temperature was 5°–−13°

For anything else, do not replace the word “to” with an en dash. Examples of constructs to avoid using the en dash include when units of measure (millimeters and meters or their symbols, names of months and days of the month for instance) appear on both sides of the word “to”; they should be fully written out:

• Wrong: wavelengths in the range 28 mm–17 m
• Wrong: wavelengths in the range 28 mm – 17 m
• Wrong: 23 July 1790–1 December 1791;    14 May–2 August 2011
• Wrong: 23 July 1790 – 1 December 1791;    14 May – 2 August 2011
• Wrong: wavelengths in the range 28 mm–17 m;   Christmas Day–New Year's Eve
• Right: wavelengths ranging from 28 millimeters to 17 meters;   from Christmas Day through New Year's Eve
• Right: from 23 July 1790 to 1 December 1791
• Right: from July 23, 1790 to December 1, 1791

---

Note that I also suggest not truncating page numbers; at least up to three digits (pp. 211–219 and not pp. 211–19), which takes care of the vast majority of books). That’s what I think is best. Greg L (talk) 15:31, 26 July 2011 (UTC)

Greg, you're not being responsive to the issue. Your "as I would do it" version has no chance of approaching consensus, since date ranges in bio are so well established, so why distract us with it? And your previous proposal to space the en dash in "28 mm – 17 m" is at odds with what I and Jeff asked for. It's OK if you disagree, and prefer the Brit style on that, but then say so; this is the area where the previous consensus was most marginal, so we're looking to fine tune it. If Jeff would make a definite proposal, that would help, too. From how this is going, I'd say finding a consensus to change it is looking unlikely. But we definitely don't want you or Jeff or feel like you haven't been heard on it. Dicklyon (talk) 15:57, 26 July 2011 (UTC)

• There’s no need to jump my bones over this. I haven’t tracked the blow-by-blow (water under the bridge) leading to where we are today and have no appreciation for what is “Brit” style or not. I only stated what I think will read most naturally for an international readership reading an English-language publication. What works for an American or Brit or an English-speaking Italian may not work fluidly when the others read it.

  So I stated my opinion as to what I think results in maximum reading flow for the widest spectrum of readers without prejudice or passion for which country’s writing customs might be disrespected (and therefore fully entitled to a road-rage brawl or “honor scuffle” in the parking lot). If expressions must be had where units of measure other than years appear on both sides of the en dash, then what is on MOS now works for me.

  I’m all turned off now over how passionate (read: borderline hostile) the tenor quickly developed here and will avoid this page until tomorrow morning. Bye. Greg L (talk) 16:11, 26 July 2011 (UTC)

  
  P.S. But, thinking about what you said regarding biographies, I amended my preference. Back to the real world until tomorrow. Greg L (talk) 16:47, 26 July 2011 (UTC)

---

For simple ranges where the en dash replaces the word “to” between two positive integers (numbers like 20 and 30), the en dash is unspaced.

• The usual amount added to Olympic-size swimming pools is 20–30 liters;   pp. 211–219;   64–75%;   the 1939–45 war

For date ranges involving years from 1000 AD or later, it is acceptable to truncate the first two digits from the second year. For all other numbers, do not truncate.

• Right: pp. 211–219;   64–75%;   the 1939–45 war;   the 1939–1945 war
• Wrong: pp. 211–19

Do not mix en dashes with prepositions like between and from.

• 450–500 people
• between 450 and 500 people, not between 450–500 people
• from 450 to 500 people, not from 450–500 people

If negative values are involved, an en dash might be confusing. Use words instead.

• −10 to 10, not −10–10
• the drop in temperature was from 5° to −13°, not the drop in temperature was 5°–−13°

For regular body text, do not replace the word “to” with an en dash if units of measure (millimeters and meters or their symbols, names of months and days of the month for instance) appear on both sides of the word “to”. Instead fully write out the expression:

• Wrong: wavelengths in the range 28 mm–17 m
• Wrong: wavelengths in the range 28 mm – 17 m
• Wrong: 23 July 1790–1 December 1791;    14 May–2 August 2011
• Wrong: 23 July 1790 – 1 December 1791;    14 May – 2 August 2011
• Wrong: wavelengths in the range 28 mm–17 m;   Christmas Day–New Year's Eve
• Right: wavelengths ranging from 28 millimeters to 17 meters;   from Christmas Day through New Year's Eve
• Right: wavelengths ranging from 28 mm to 17 m
• Right: from 23 July 1790 to 1 December 1791
• Right: from July 23, 1790 to December 1, 1791

For the first sentence of biographies, citations, and other sections of articles where parenthetical-like data is customary and expected, the word “to” may be replaced with an en dash in a ranges of dates. When doing so, place a space on both sides of the en dash:

• Right: 23 July 1790 – 1 December 1791
• Right: July 23, 1790 – December 1, 1791
• Wrong: 23 July 1790–1 December 1791

---

Greg L (talk) 16:47, 26 July 2011 (UTC)

I didn't know you were such a sensitive guy, Greg, so please excuse my clumsy nudge. I just wanted to point out that there's some delicacy involved if we're to fine tune the current consensus to a better one. Jeff is the one can tell you how much opinions are worth (his "I just like it" complaint), so I thought we could focus on where they differ, and talk logically about those differences; you don't need to know the gory history, but need to recognize that I was trying to offer a step toward Jeff's logically-preferred position (or is it just his opinion?); and you stepped back from there while saying you were trying to represent our emerging consensus (at least on the 28 mm – 17 m thing).

Here's what you need to know: on this spacing issue, the old MOS (stable for many years before recent months) reflecting the Brit view was crafted by Tony and Noetica and others, in a good discussion circa 2007 or earlier. It calls for spaced en dashes more often than was natural with us Yankies. When we had the big vote on the details, this area came out quite mixed, even after Tony and Noetica agreed to compromise and use the spaces only in more complicated things like date ranges. I was happy. A majority was happy. Jeff wasn't happy, because he wanted to see it move it bit more, and because he didn't like the fact that we didn't quite find a strong consensus. PMAnderson made no comment on these provisions until we were all done, and then used it say that consensus was not achieved, so we shouldn't accept the new en dash section as a whole (it's still not clear if he wants to see changes in this section, or whether he likes it but just doesn't want to see anyone claiming the the MOS has consensus).

So I wanted to give Jeff a chance to help find a better consensus. But Jeff has continued to decline to propose a version he likes, and you've not helped with your "support" of my proposal, so I think I'll give up. I don't really care if we keep the version you boldly changed to, or go back to what Noetica had done. My impression is that we are close enough to "consensus", meaning there's nobody who can't live with it, unless someone else is willing to say exactly how we need to change it make them happy. I don't think anyone but Jeff is likely to give that a try, but we'll see.

And I don't think any of us are getting passionate, personal, or otherwise steamed up over it—I'm sorry if my text gave that impression. Dicklyon (talk) 21:50, 26 July 2011 (UTC)

Um, Greg, what are you waiting for me to propose? I’d have been glad to do it had I realized you were waiting :-). My approach may be slightly different from that of others—I usually look to first discuss concepts (like closed-up dashes for ranges of physical quantities). Based on years of experience with committees, working from a document “to argue away from” usually ensures much argument (perhaps not so much as here, but that could simply be because fewer people are involved). How often have I prevailed in getting agreement in concept before drafting? Far less often than I would like. What has it cost me? In some cases, several years and suboptimal results.

Is my preferred practice more “logical”? That’s impossible to resolve—I think it’s really a matter of “Iwannadoitmyway”, as many here have mentioned (Darkfrog24 most recently). I can certainly give reasons why I think some aspects of “North American” practice are better, but Noetica and Tony might not agree. But I again note one thing: however it’s determined, North Americans are a substantial fraction of the English-speaking and English-publishing world—and for the most part, NA style guides seem to lead the sales in the US, Canada, and the UK. By no means does this mean that NA practice is any better than any other—CUP is hardly run by a bunch of idiots—but it’s not completely nuts, either. And I think it has sometimes been given short shrift here.

I certainly wasn’t looking to take you on, and I doubt that Dick was either. Sorry if I came across that way.

In any event, let me know what I haven’t produced and I’ll try to better describe it. JeffConrad (talk) 01:17, 27 July 2011 (UTC)

Jeff, we're tired of concepts and just want to settle on a version; I applaud Greg for boldly pushing for what he thought would work, even if I thought it was unlikely to please you (and though I pushed back on some of his intermediate versions). If you like Greg's, say so. If not, propose a specific change or a new version. Noetica asked you several times for a definite proposal, and got nothing but long discourse and complaints about process and his use of "we". This doesn't need to be as slow as a standards committee. Excuse my increasing grumpiness. Dicklyon (talk) 02:27, 27 July 2011 (UTC)

Jeff Conrad’s proposals

Dick, for practical purposes, I actually did submit a proposal. I normally follow North American practice, which has very simple rules: en dashes are always unspaced, so few examples are needed. We seem to need such an extensive treatment because our rules are so complicated. Now NA guide aren’t necessarily right, but they do far outsell all others, in some case even outside NA, so they can’t be all that bad, either. And above all, they’re simple. Note though, that in polling, I was OK with spacing dashes in ranges of full dates, largely recognizing Tony’s comment that the style is well established here, especially in bios. And I also don’t care for 23 July 1790–1 December 1791 because I initially see “1790–1791”.

Again the rules are simple:

---

Approach 2

Dashes are spaced when both endpoints are full dates (e.g., some combination of numerical date, spelled-out or abbreviated month, and year):

• 23 July 1790 – 1 December 1791 not 23 July 1790–1 December 1791
• June 25, 1988 – April 3, 1989 not June 25, 1988–April 3, 1989

---

We’re done . . .

This didn’t seem to be what most folks wanted; allowing for other personal preferences, I might add

---

Approach 3

Dashes may be spaced in other ranges of dates for which both endpoints contain a space:

• 24 June–23 July 2010 or 24 June – 23 July 2010
• May 1–December 3 or May 1 – December 3
• Christmas Day–New Year's Eve or Christmas Day – New Year's Eve
• Christmas 2001–Easter 2002 or Christmas 2001 – Easter 2002

Dashes are always unspaced in ranges of dates, times, physical quantities, or similar, for which only one endpoint contains a space:

• 17–22 April not 17 – 22 April
• August 6–8 not August 6 – 8
• 5:45–6:30 p.m. not 5:45 – 6:30 p.m.
• 4–20 mA not 4 – 20 mA
• 24–105 mm not 24 – 105 mm

Dashes are always spaced in ranges of time that span more than one day:

• 10:30 pm Tuesday – 1:25 am Wednesday not 10:30 pm Tuesday–1:25 am Wednesday

Dashes may be spaced in other ranges of time for which both endpoints contain spaces:

• 10:30 pm–1:25 am or 10:30 pm – 1:25 am

Dashes may be spaced in other ranges for which one or both endpoints are long or complex, and for which closed-up use could be confusing; it is impossible to cover every case, so sometimes the editor must exercise judgment.

Dashes may be spaced in ranges of physical quantities for which both endpoints contain spaces, as when units are repeated for both endpoints or when the endpoints have different units:

• 28 mm–70 mm or 28 mm – 70 mm
• wavelengths in the range 28 mm–17 m or wavelengths in the range 28 mm – 17 m

Care should be taken when spacing en dashes in ranges of physical quantities because of possible confusion with the minus (−). Confusion seldom arises with ranges of dates because they are usually not subtracted; however, subtraction of physical quantities is a common occurrence. Because the minus is normally spaced, confusion can sometimes be avoided by closing up the dash:

• 20 Hz–20 kHz

Quantities with units or unit symbols on both sides of the dash are often more readable when written out using to, especially when the units are spelled out:

• wavelengths in the range 28 mm–17 m; sometimes better, wavelengths ranging from 28 mm to 17 m
• 20 Hz–20 kHz; sometimes better, 20 Hz to 20 kHz
• wavelengths in the range 28 millimeters – 17 meters; better, wavelengths ranging from 28 millimeters to 17 meters

Recasting may not be practical when space is at a premium (as in a table), or when the use is adjectival:

• 20 Hz–20 kHz response not 20 Hz to 20 kHz response or 20 Hz-to-20 kHz response

Complex ranges of dates are also often better recast:

• Christmas Day–New Year's Eve or Christmas Day – New Year's Eve; better, Christmas Day through New Year's Eve
• from 23 July 1790 – 1 December 1791; better, from 23 July 1790 to 1 December 1791

---

To be honest, I’m not sure I really prefer the recast of “Christmas Day–New Year’s Eve” to “Christmas Day through New Year’s Eve”. And I′m still not convinced that this is a date (and even if it is, why is confusion more likely than with “San Francisco–Denver flight”?) In any event, I’ve listed it as a date so as not to get sidetracked.

Note that I don’t call for barring using less than the full range of page numbers. I usually include the last two (e.g., pp. 1137–39, pp. 1407–22). Obviously, this doesn’t work if the first number in the range changes: pp. 883–901. Comprehensive guidance on this obviously would require some additional work.

At least to me, most of this isn’t too different from what several others have suggested.

I hardly suggest that this is sufficiently polished to incorporate as is (though perhaps it’s close). In particular, additional spacing between examples on the same line would probably improve things. And I’ve probably overlooked something. But this should give a general idea. In essence, I would allow either North American or British practice (add scare quotes if you must). In that sense, at least philosophically, I’d recommend everything, but the biggest initial issue is with spaced dashes in ranges of quantities—I don’t like ’em.

With the optional usages, most of my preferences (unspaced) are at the left. Are they better than the ones with spaces? There’s no answer. But are they worse? If so, I’d like to know why. And if so, why are they acceptable to (and often called for) the majority of widely used guides?

Greg, I hope this addresses what you were requesting, whether or not you agree with my ideas. JeffConrad (talk) 07:15, 27 July 2011 (UTC)

OK, better; a too-long and too-short proposal, but now we know more specifically what you're thinking. I hadn't realized that you wanted to leave such a wide range of cases up to editor discretion; if we do that, we can do it with many fewer examples, don't you think? But I felt like there was stronger support for more definite guidance, than for more options (there was some of that, too, I recognize). Let's see what Greg thinks, and whether anyone else cares. I can draft an intermediate-length polished version if it looks like that's where the wind is blowing. Dicklyon (talk) 19:12, 27 July 2011 (UTC)

• Whatever you guys want. I’m tired of so much fuss on en dashes. It should be less than 1% as difficult as getting dates delinked and ridding us of the IEC prefixes but has exceeded that threshold. Greg L (talk) 19:40, 27 July 2011 (UTC)

Dick, easier to trim a long one that to guess at what I want from a short one, right? On editor discretion: my real preference is pretty simple—I would prefer to do things in the manner to which I am accustomed, perhaps with the exception of ranges of full dates and times. But we know that clearly would not fly. I, like several others, have objected to a seeming imposition of what’s arguably minority practice, and if Butcher is the guide, one that is permissive rather than mandatory. We could argue all day (Day? Who am I kidding . . .) about which is better—some could cite “five major guides” that support the more frequent spacing, and I could cite ten arguably “more major” guides that call for everything to be closed up. I think we know where that would lead . . . So unless we battle to the death over forcing people to follow something to which they may object (with either British or NA practice), allowing either seems the only way. And quite honestly, we seem to be in pretty good agreement about allowing spaced en dashes for parenthetical dashes is essentially allowing either non-OUP British practice or NA practice, and and in doing so acknowledging the either is valid. I have a fair number of books from quality British publishers, and honestly find the spaced en dash a bit distracting—but I suppose many accustomed to the British style have the same reaction to unspaced em dashes. And the same is probably true of unspaced vs. spaced en dashes. But it’s worth noting that the usages that we’ve been discussing to death are pretty infrequent—I have to search mighty hard through all of my CUP and Elsevier books to find even a few examples. More succinctly: we seem to have no problem with allowing two different styles for a usage that is very common but get all worked up about doing so for a usage that is comparatively rare.

I’m sure someone will take issue with my characterization as “British” or “North American”. But Chicago, OUP, Garner, and others describe it as such, so what’s the big deal? Compared with AmE–BrE differences in spelling, these differences seem pretty minor.

So I guess I proposed what I did because it’s probably as unreasonable to insist that people do it my way as I claim it is for others to insist that I do it their way. I agree with you that many editors seem to have wanted more specific guidance. But if we insist on a more prescriptive approach, we are faced with which one to prescribe; I would of course insist on the one I prefer, and for which I could make a pretty strong case is very much the majority approach. And the the debate would continue forever . . .

As for the length: a bit of rearrangement and consolidation might slightly reduce the size. I was so tired when I wrote it that I could barely stay awake . . . I’m amazed it’s even semi-coherent. JeffConrad (talk) 21:32, 27 July 2011 (UTC)

Greg, you objected that I had not made a proposal, and you now almost seem to be objecting that I have. Am I missing something? For what it’s worth, having recently been through a long and pointless discussion about quotation marks, I wanted nothing to do with this discussion, as Dick can attest. I completely agree with you that the effort expended here boggles the mind, especially when I think of what could have been accomplished had the same effort been expended on weightier issues. And there’s still significant disagreement on a couple of points . . . I’ve been using en dashes for 25 years—almost as long as I have had access to something that could render them. For guidance, I simply grabbed a couple of widely used guides (primarily CMOS), spent a few minutes reading what was recommended, and moved on. Here, of course, not everyone prays to the same Guide, with each side insisting that “my Guide is bigger than yours”. And until we can resolve which is the One True Guide, the debate will continue. We could, of course, recognize that there may be more than one True Guide, which is something like I have proposed. JeffConrad (talk) 21:51, 27 July 2011 (UTC)

No, I’m not objecting to anyone’s proposals, Jeff. Everyone active on this issue as of late is knowledgeable, acting in good faith, and is plenty civil enough. I just don’t understand why so much effort has been (and currently is being) spent on en-dash stuff. I’ve grown weary of it because I see so much effort being expended and the car is still spinning its wheels in wet snow. I’ve thrown out my own ideas so those of you who are still keenly interested in further fleshing this stuff out can adopt bits & pieces of it, the whole thing, or nothing at all.

I personally think that en-dashes should be kept to a minimum because the word “to” is really easy to read. So all this head scratching trying to figure out how to expand the circumstances under which we can replace something über-easy to read with an en-dash and still have the text be easily parsed by the eye seems a brain damaged endeavor.

I lump the expanded superset of en-dash usage in the same camp as single-divisor math featuring negative superscripted exponents like A rocket engine with a flow rate of 300 L s^–1 rather than the far more accessible A rocket engine with a flow rate of 300 liters per second or even A rocket engine with a flow rate of 300 L/s. Oh, yes; we all know there is far too much of this going on all over Wikipedia. I’m certain there are many editors who think they “appear smart” because they “write smart and abstrusely and all *scientificy*.”

Which brings up another flaw; Wikipedia is far, far too quick to launch into unit symbolitis rather than spell them out. Most quality publications would write A first-press squeeze of 100 kilograms of olives would produce 20–25 liters of extra-virgin olive oil. We all know there are many wikipedians, who deep down think that using unit symbols is futuristic and cool, who would write A first-press squeeze of 100 kg of olives would produce 20–25 L of extra-virgin olive oil.

That’s my pout for the day. Thank you for listening. Greg L (talk) 00:19, 28 July 2011 (UTC)

Ahem... ― A. di M.​_plé​^dréachtaí 00:58, 28 July 2011 (UTC)

Oh, well then… “Fuck” is 33 times more common than “kg”. That doesn’t tell us what is good practices for regular body text in an encyclopedia, does it?

• Wrong: 24 June–23 July 2010
• Correct: Oh, fuck it.

I’ve got statistics on my side. Greg L (talk) 02:18, 28 July 2011 (UTC)

A brief digression on units and the clarity thereof

Though I think superscripted negative powers are a bit hifalutin for most contexts, I still may be one of those evil trendies who generally prefer unit symbols—but it all depends on the context and the frequency of use. And the complexity of what is to be spelled out. A good example can be found in numerous stories about the San Bruno, California, gas pipeline explosion. For the most part, the normal operating condition was correctly described as “a pressure of 375 pounds per square inch”. But this gets pretty tedious if it’s mentioned several times in an article, and this apparently inspired some pretty creative writing, including “375 pounds per square inch of pressure”, and even “375 pounds of pressure per square inch” (lbf/in⁴?). Clearly, the first form is the only one that is unambiguous; perhaps the second could be handled as “375 pounds-per-square-inch of pressure”, though this seems awkward to me. The adjectival form might be “375-pounds-per-square-inch pressure”. The unit symbol form definitely has appeal here, but even so there are problems: “lb/in⁴” is ambiguous, because lb more often refers to mass than to force; “lbf/in⁴” isn’t ideal, either, because lbf isn’t standard symbol, and most non-technical readers don’t know the difference between pounds mass and pounds force anyway. Here I might prefer something like “a pressure of 375 pounds per square inch (lb/in²)”, with the symbols used thereafter or perhaps even “a pressure of 375 pounds per square inch (psi)” despite the universal deprecation of psi (which is probably the term most American engineers would use in informal conversation). Oh, f—k it . . . There should be no doubt that I consider avoiding ambiguity with physical quantities mighty important, but in some cases there just isn’t an obvious best approach.

Getting back to topic: I think most of us would agree that clarity and easy reading are paramount. But we apparently don’t always agree on the forms that are easiest to read—probably has a lot to do with the forms to which we are accustomed. JeffConrad (talk) 08:14, 28 July 2011 (UTC)

When a unit of measure is long, like the pipeline was pressurized to 375 pounds per square inch, and is repeated many times in the article, my advise for good technical writers is to parenthetically give the unit symbol the pipeline was pressurized to 375 pounds per square inch (psi) and to then rely only upon unit symbols (after the explosion, older pipelines had their operating pressure reduced to 290 psi).

Depending on the nature of the article (if it is particularly long and prone to readers skipping to particular sections of interest, or if the parenthetical introduction to the unit symbol “psi” is mentioned early in the article and then doesn’t crop up again until way later), I will repeat the spelled-out version and the parenthetical introduction to the unit symbol when the unit of measure is once again mentioned: it was later certified for operation at up to 350 pounds per square inch (psi).

As technically minded writers, we get used to this stuff. As a mechanical and R&D engineer, I can authoritatively tell you than many people—typically women—are not as familiar with “psi” or pressures of any sort. Plain-speak and reinforcement of what unit symbols mean is important and is a lesson lost on many contributors who are intimately familiar with the subject matter. European writers tend to think that everyone including the extra-smart German Shepard next door knows what “kg” means. There are still plenty of people in America who were adults during the Korean war who get quickly lost when they see 100 kg of olives. That is so needless; kilogram doesn’t take up that much room and the vast majority of articles wouldn’t mention the unit so terribly often that such a simple, single-word unit of measure like kilogram would bog the article down in eye-fatiguing tedium.

We all have to bear in mind that Wikipedia is directed to a general interest readership from all walks of life. I think editors who write this sort of stuff: the typical garden hose in America delivers only about 0.5 L s^‒1, ought to be required to pass a technical writing skills test before being allowed back. Greg L (talk) 18:10, 28 July 2011 (UTC)

Well, to those particular people writing 100 kilograms of olives wouldn't be particularly useful, either. 100 kilograms (220 lb) of olives is better, but in some particular cases 100 kg (220 lb) of olives is preferable. ― A. di M.​_plé​^dréachtaí 23:30, 30 July 2011 (UTC)

@JeffConrad: I fully agree. Also, while empirically estimating which form is easier for readers to read is quite non-trivial (there are a few studies about that, but the sample sizes aren't huge and the participants are usually undergraduate students, who might not be representative of the population as a whole), estimating which one they're most accustomed to is much easier (there are several language corpora available on the Web for free). ― A. di M.​_plé​^dréachtaí 23:30, 30 July 2011 (UTC)

New subpage to work on en-dash spacing

Wikipedia talk:Manual of Style/endash spacing is where I've put a new proposal and discussion to keep from cluttering this page where lots of people are weary of en dashes. Please join if you care. I'm trying to make a version that's not so long and makes Jeff happy by allowing more options, and then we can see if that can raise a consensus better that what we had before. Dicklyon (talk) 05:49, 29 July 2011

If this were Dicklyon's goal, it would have been simple. But, oddly enough, the same text has reappeared, and and being reverted for, alothough it was sidely opposed at the original poll, is not supported by sources, and is opposed at the discussion. Funny thing that.Septentrionalis PMAnderson 00:36, 2 August 2011 (UTC)

Using endashes in compounds whose elements themselves contain hyphens

The section on endashes used to contain a case for using endashes in compounds whose elements themselves contain hyphens or spaces (e.g. the anti-conscription–pro-conscription debate) and when prefixing an element containing a space (pre–World War II technologies, ex–prime minister). I'm a little late to the game and I see that there has recently been a long and protracted debate on rewriting this section of the style guidelines. I've tried to follow the discussion as best as I can, but it's quite long. I discovered a draft page where various elements of the draft were voted upon, and I see that this particular case received wide support (see here). It received 27 votes in support, 2 opposing votes, and 1 neutral vote. Yet, despite this support it somehow didn't make it into the final draft, and I can't find the discussion where it was axed. Can anyone shed some light on why this case doesn't currently exist in the guideline? —SW— ^yak 22:02, 29 July 2011 (UTC)

It got reorganized some, but those examples are still there, more or less (search for "ex–prime minister" or "a pro-establishment–anti-intellectual alliance"). The last bit is a simple application of a general use, the hyphens being irrelevant. Nobody came up with an example where having hyphens in one part or the other was a good reason to use an en dash. Dicklyon (talk) 22:23, 29 July 2011 (UTC)

Ah, I see you're into this proposal: Wikipedia talk:Naming conventions (ships)#Punctuation and ship classes. Example would be "Chungmugong Yi Sun-shin–class destroyer" versus "Chungmugong Yi Sun-shin-class destroyer". Good question. We ended up taking out the guideline to use the en dash on account of the spaces in Chungmugong Yi Sun-shin, and the one with the hyphen pretty much went with it, for lack of a compelling example of where it would be useful. Should we reconsider? This remains an area of less than perfect concensus, but it would be some work to reopen it and try to do better, as I'm trying to do with the spacing issues. Dicklyon (talk) 22:36, 29 July 2011 (UTC)

There are a couple dozen ship class names that have hyphens in the name itself. See a list I compiled at User:Snottywong/Ship classes/Pre-hyphenated. I don't have any data on how many non-ship-class articles might be in the same boat (no pun intended). Take a look and let me know if you think it merits some more work. —SW— ^chat 23:11, 29 July 2011 (UTC)

A hyphen in such constructions works fine: Duguay-Trouin-class cruiser (much better than a space, which suggests you're talking about some kind of a class cruiser). Dicklyon (talk) 00:04, 30 July 2011 (UTC)

• The only function of a dash is this context is to prevent the confusion possible if the same pubcation is used at two different levels of grouping; Lloyd-George–Winston-Churchill Government to avoid suggesting four men, Lloyd, George, Winston, and Churchill, instead of the Giovernment led by Lloyd George and Winston Churchill. This implies the diminishing, but still attested Britsh custom of hyphenating compound adjectives; but Government led by David Lloyd George and Winston Churchill is preferable anyway. Septentrionalis PMAnderson 23:26, 29 July 2011 (UTC)
• Since Chungmugong Yi Sun-shin-class destroyer is unambiguous, it needs no dash; the structure is shown by the italics. If there is any danger of ambiguity, it rasts in the dash, which might suggest a compound between the Chungmugong Yi Sun-shin and "class destroyer." Septentrionalis PMAnderson 23:26, 29 July 2011 (UTC)
• The rather more common U-10 class submarine doesn't really need any more punctuation; it's clear as it stands. Does anybody add dashes or hyphens to it? Septentrionalis PMAnderson 23:43, 29 July 2011 (UTC)

Some writers use en dashes when compounding hyphenated words, but they're relatively uncommon; I've come across quite a few style guides that recommend against the practice, with the argument that it rarely dabs anything, and tends to disrupt the attention of the reader, a general no-no for punctuation. But en dashes are common with spaced compounds. (For Anderson's comments, read 'English' where he says 'British', 'increasing' where he says 'diminishing', 'one' where he says 'the only', and 'attributives' where he says 'adjectives'.) However, since the italics make the class name obvious in Chungmugong Yi Sun-shin-class destroyer, a simple hyphen is sufficient, assuming that we ignore those who are protesting against italics.

As for U-10-class submarine, again the italics, assuming they're correct, are sufficient. Where italics are not used, I don't know. For simple names, the Armed Forces journal international uses hyphens (N-class submarine) whereas Jane's fighting ships uses scare quotes ("N" class submarine). The latter convention would of course handle compound names as easily as simple ones, but it has not generally been what we've used on WP. Even without italics, though, I'm not sure using an en dash with a hyphenated class name provides any benefit.

Where the en dash would be useful is in things like Passenger-Only Fast Ferry–class ferry and P 4–class torpedo boat. — kwami (talk) 00:34, 30 July 2011 (UTC)

Kwami may choose to read what is not the case; but I see no evidence that Scottish usage varies from English, British usage of the hyphen has been diminishing for over a century (although it may have hit bottom), and attributive (as a noun) is jargon. What other function of the compounding dash is there other than clarity? Power-gaming is not a function, it's an abuse. Septentrionalis PMAnderson 14:58, 30 July 2011 (UTC)

Other meaning of "English": not of England, but of the English language. I know, when you're not claiming this "isn't English", you're claiming it's "only British". Perhaps you could invent some new excuses? Though I see promise for a new one: those you disagree with you could now be "abusive". Disappointingly unimaginative, though. — kwami (talk) 19:28, 30 July 2011 (UTC)

I′ll point to an example that Dick Lyon pulled out a while back: San Francisco–based company. Now perhaps the caps would help avoid ambiguity if a hyphen were used rather than a dash, but I think they help less here than in post–Civil War period because the caps don’t directly abut the hyphen that is to be replaced. The example quasi–open government policy from CGEU doesn’t even have any caps, and quasi-open government policy leaves me with a slightly different impression.

I agree with Chicago that many readers will miss the subtle distinction here. But in any event, if some readers just see San Francisco–based company as San Francisco-based company, what has been lost by using the dash? I do agree that a compound can only be so long for this to work, and the CGEU example is getting close, as is pre—World War II aircraft. JeffConrad (talk) 22:37, 30 July 2011 (UTC)

Yes, really long strings screw up all hyphenation conventions. As for post–Civil War period, some writers capitalize the "Post", in which case capitalization does not help. — kwami (talk) 02:25, 31 July 2011 (UTC)

Unit measures for dates

What are "unit measures for dates"? Two ounces of July 30? 23rd century – 4th millennium? Spring – summer? Last weekend – next Tuesday? Last fortnight – next month? Unless I'm the only guy who doesn't get it, it should be re-explained. Art LaPella (talk) 22:29, 30 July 2011 (UTC)

Well, there are at least two of us . . . and I’m pretty familiar with units. JeffConrad (talk) 06:15, 31 July 2011 (UTC)

Template:Resolved by re-explaining. Art LaPella (talk) 16:50, 31 July 2011 (UTC)

Greg's edits are not at all "resolved". They depart significantly from the consensus version we spent months negotiating, without consensus. At this stage, only uncontroversial alterations should be made. Tony (talk) 06:13, 1 August 2011 (UTC)

I think Greg's version is not bad, now that the "units" thing has been fixed. I also agree with Tony that "only uncontroversial alterations should be made." So I was fully expecting Greg's version to get reverted back to the consensus version by Noetica, by someone who felt that Greg moved it in a bad direction; so far, nobody has complained or reverted. So maybe it's not controversial? Tony, are you saying you don't like it as much as what we had before? If that's what you're saying, then by all means let's take it back to the consensus version; I've invested a lot of effort trying to find a version that nobody objects to (see Wikipedia talk:Manual of Style/endash spacing), but if it doesn't exist, then the one that the majority at least was OK with should be OK for now. Dicklyon (talk) 06:32, 1 August 2011 (UTC)

This particular issue this thread is about was solved. The other ones are being discussed at /endash_spacing. (As for me, I prefer Noetica's version as it's shorter and clearer, though I'd retain Greg L's suggestion to avoid über-complicated dashed ranges.) ― A. di M.​_plé​^dréachtaí 13:54, 1 August 2011 (UTC)

I consider the discussion at /endash_spacing to be finished, and failed; I've reverted to Noetica's version, the last one that had something like consensus. Dicklyon (talk) 17:16, 1 August 2011 (UTC)

Asterisks in examples

Weren’t these added to assist readers who are red–green colorblind? I agree that the asterisks don’t look great, but some accommodation seems reasonable. If the asterisks are too unsightly, what about something like !Another "planet" was detected — but it was later found to be a moon of Saturn, in keeping with the symbol used to distinguish the templates? Although the implication of bang may not be obvious to everyone, it’s nonetheless common in computer science, so the meaning should be more obvious than using an asterisk. In any event, a brief mention of the convention should clarify. JeffConrad (talk) 23:37, 25 July 2011 (UTC)

Whatever we do, we don't need a special notation for the en dash section. I think Noetica made up the asterisk for people reviewing the new draft, but didn't expect it to last; so now is the time to work something out. The "!" might be OK; but can we fix the template to do it automatically? Or just adjust the colors to be distinguishable enough? Dicklyon (talk) 02:59, 26 July 2011 (UTC)

I strongly prefer we just use the words [correct] and [incorrect] instead of any color change and symbols, so no need to worry about visually impaired reader's accessibility. -- Sameboat - 同舟 (talk) 04:18, 26 July 2011 (UTC)

I agree that we don’t need (or want) a special version for the dash section, and if it were decided that some indication other than color is indicated, building it into the template might be the only way to ensure consistency. Of course, this well could raise objection from some who don’t want any such clutter. But there may be even more merit to Sameboat’s suggestion, which doesn’t rely on color perception or a gimmick like a “!”. When both “correct” and “incorrect” examples are on the same line, putting not between them as A. di M. did works just fine and is shorter. Both approaches are consistent with how most other style guides deal with it. I see no problem with retaining using color for the examples to make it clear that they aren’t part of the text.

Another approach would be to go even further and surround the example with curly braces, as CMOS and Garner do (e.g., {Another "planet" was detected — but it was later found to be a moon of Saturn}) so that it’s more obvious to the visually impaired. The typeface change certainly helps, but it’s nothing like the change of color and typeface. If such a change is made, it should be incorporated into the template to avoid the messy construction I had to use here.

Probably not the top priority, but perhaps worth thinking about. JeffConrad (talk) 07:39, 26 July 2011 (UTC)

The bang might be standard in computer science, but the asterisk is standard in grammar (though usually for genuinely unidiomatic or unattested usages, not merely ones which are the wrong register) so I wouldn't expect either to be much clearer than the other. I once proposed this, but it was opposed as too cluttered. Anyway, whatever scheme of colours/symbols is used should be redundant with the surrounding context, not relied upon. ― A. di M.​_plé​^dréachtaí 12:44, 26 July 2011 (UTC)

I argued strongly and at length on Noetica's Talk page in favor of using "Correct:" and "Incorrect:" (see Sameboat's note above), with appropriate font color and typeface contrasts, because the correct/incorrect convention is both common in other usage manuals and guides, as well as in other sections in WP:MOS. Noetica agreed and assured me that he had "made up the asterisk for people reviewing the new draft, but didn't expect it to last," as Dicklyon noted above.

• That revision is the "minor" format change Noetica is referring to in parentheses in the list (far) above of those who supported posting the draft and moving on to the next stage in review and approval, as I did, contingent on making this revision ASAP after posting the draft. His priority was getting the draft posted for comment without further delay because it's quite behind schedule, and that seemed to me to be a reasonable course of action.
• I also argued for including a brief, 1-sentence explanation of the asterisk convention and of its interim use to avoid exactly this kind of confusion. As JeffConrad says above, "In any event, a brief mention of the convention should clarify." But evidently Noetica either didn't consider that to be necessary or overlooked it in the process of posting the draft.
• As for using the "bang" instead of an asterisk, with all due respect to JeffConrad (whose numerous valuable contributions are noteworthy), it seems to me to be a non sequitur to assume that because the bang is "common" in comp sci, ergo it's "meaning should be more obvious than using an asterisk." (It wouldn't be to me, and I was a comp sci minor.) Aren't relatively few MoS users likely to be experienced in programming languages that use a bang extensively? Besides, ! is a very narrow symbol, so it's less prominent than many other common special characters, such as #. (* is also obviously quite small and is quite indistinct when it adjoins the 1st letter of what follows, without an intervening space. But I can imagine now the fierce debate that would ensue over how large a space is best to use!)
• Regarding the contrasting font colors and typefaces, they may perhaps improve readability when the correct and incorrect forms appear imbedded in-line in the text vs. being set off on 2 separate lines. (As discussed elsewhere, color differences alone wouldn't be visible to those with impaired color vision, and the color sets used by some browsers and computer models often render color contrasts less distinct, even under the best of circumstances.) I concur with A. di M. that "whatever scheme of colours/symbols is used should be redundant with the surrounding context, not relied upon."

• But we should not downplay the importance of paying strict attention to readability: Parts of the MoS "Hyphens" section provide an especially egregious example of how indistinct the contrast of the current red-green serifed typeface can be. One portion of that section literally (and I do mean literally) fills my laptop's 15" screen from top to bottom and side to side with single-spaced text and very little white space! The red-green serifed typeface is barely noticeable unless I look carefully at each line and read more slowly than I normally would. It's a textbook example of poor page design! And it's likely to discourage editors from paying much attention to that section of the MoS if they can avoid it. The MoS can, and should, do better, and that change could begin with us. ("If not us, who? If not now, when?" etc.) — Jack --Jackftwist (talk) 22:21, 26 July 2011 (UTC)

The problem with the color change is that it changes the font-color, since the color is so dark, it is often indistinct from the non-colored black text. However a pale colored background should solve this problem nicely. This is my proposed change to the current format:

[incorrect] {This are the wrong example.}

[incorrect] {<span style="background:#fdd; font-family: Georgia, serif;">This are the wrong example.</span>} 

-- Sameboat - 同舟 (talk) 01:31, 27 July 2011 (UTC)

(after ec) Bang! Bang! Bang! OK, OK, I was just sayin’ . . . I thought I had made it clear that using correct/incorrect (or right/wrong, which is shorter and monosyllabic) is better, and A. di M.’s not is the briefest and probably the best for columnar examples. In any event, the color difference is at best a minor convenience—I’m fine with keeping it, but agree with many others that we should not rely on it.

Just out of curiosity: what about my suggestion to use something like braces, as in {pseudo–page transition}, to make the examples stand out better when color and typeface alone aren’t sufficient, for whatever reason? The braces aren’t especially pretty, but then neither is almost any convention I’ve seen for literal examples (especially in computer manuals, even the good ones).

I tend to agree with Sameboat that the color change can be hard to see, but the pink background seems a bit garish, especially if I had to look at a page full of it. JeffConrad (talk) 01:42, 27 July 2011 (UTC)

I like both Sameboat's and JeffConrad's suggestions. We could consider alternatives of which background color and symbols to use and discuss them.

• Jeff, like you, I like the brevity of right/wrong, but it no doubt violates some WP policy for being too blunt. And IMO {braces} aren't especially any more or less attractive than some of the alternatives. Maybe square brackets in a larger font than normal text (to avoid confusing them with intentional punctuation) might have more esthetic appeal, or we might be able to find a symbol in LaTex that would be more visually pleasing.
• As I understand it, the purpose of whatever symbol is used (e.g., braces, asterisks, etc.) is primarily for the benefit of users with impaired color vision. Adding a background color, in combination with the symbol, has at least 2 advantages:

(1) It would make the incorrect example stand out when both the correct and incorrect forms are shown in one line of text. No combination of symbols with (or without) colored, serifed fonts does that particularly well, especially where a lot of examples are shown in-line, as they are in the "Dashes" "Hyphens" section I referred to in my previous comment. They simply get drowned out by the sea of regular text.

(2) Yes, a page full of pink might be garish, but it would sure get the reader's attention (the whole point), and its reddish character virtually screams "Stop!" But seriously, there's much more flexibility in using background colors than font colors. Font colors must be dark, which creates the problem Sameboat and others have mentioned of becoming indistinct from normal black text. But a much wider variety of light or medium background colors could be used—perhaps a soothing pastel blue or yellow? Or amber? (Green might create confusion because it often connotes "Go!" or "OK".)

In any case, this merits further consideration. — Jack --Jackftwist (talk) 17:23, 27 July 2011 (UTC)

Some adjustments as per suggestions above: [Placeholder.] (Right). [Placeholder.] (Wrong)

<span style="font-size:150%">[</span><span style="background:#efe; font-family: Georgia, serif;">Placeholder.</span><span style="font-size:150%">]</span> <span style="font-size:85%">(Right)</span>

The right/wrong are placed behind the examples so we don't need to concern about the capitalization when making it into a template. -- Sameboat - 同舟 (talk) 01:12, 28 July 2011 (UTC)

At least to me, the pastel backgrounds are pretty garish—far more so than the current scheme. I certainly would not want to read very much of this. I also think the curly braces work much better than the enlarged square brackets because they don’t increase the line spacing with inline examples. The curly braces are sufficiently uncommon in normal text that confusion is unlikely to result; moreover, there is some precedent for the use of braces, so the reader is more likely to be familiar with them. But the braces are jarring nonetheless, so if this idea were to gain any traction, I would limit use to inline examples—columnar right/wrong examples (and similar) are usually obvious enough that readers probably don’t need the extra cues. Anything that disrupts the normal flow is distracting, and I think it should be kept to a minimum—including typeface changes that aren’t central to distinguishing example text from ordinary text. I like to tell myself that I had no choice here, but sometimes I wonder—suffice it to say this is painful to proofread. JeffConrad (talk) 02:10, 28 July 2011 (UTC)

I surely hope none of you are serious about proposing these obtrusive forms. The current way of using the templates {{xt}} and {{!xt}} together with a variation of approving or rejecting additions is close to perfect. −Woodstone (talk) 14:56, 28 July 2011 (UTC)

Then I prefer to not change the font style of the example text at all. You know what? The color and font changes make me glare at the screen for more than 1 second to distinguish the example from other text. It is absolutely flawed. The easiest and non-obtrusive way to indicate the right/wrong example is simply to italicize the example text (which is how dictionary does) and precede it by green tick or red cross with alt attribute and title attribute. -- Sameboat - 同舟 (talk) 16:07, 28 July 2011 (UTC)

Re Woodstone: And I, in turn, sure hope you aren't serious that the "current way of using the templates {{xt}} and {{!xt}} together with a variation of approving or rejecting additions is close to perfect." [Italics added for emphasis.]

• First, "close to perfect" assumes some objective criterion or set of criteria for determining what is perfect. E.g., under U.S. baseball's definition of a perfect game, a game in which the pitcher(s) allowed only a single walked batter to reach base, and no one else, could reasonably be described as "close to perfect." So would achieving a 99% score out of a maximum possible 100% on an examination (i.e., no "bonus points"). Are there any comparable criteria for defining a perfect way of contrasting right and wrong examples so one is clearly discernable from the other, etc.?
• Second, the noteworthy disadvantages of the {{xt}} and {{!xt}} templates have already been discussed at length above, in particular that the font colors must be dark, so they're often not easily distinguishable from the surrounding black text, especially in in-line examples that are imbedded in a long passage, so that many lines of text cover entire lines on the screen. (Again, see the MoS section on "Hyphens" for an egregious example of how difficult it is to readily identify the examples using the {{xt}} and {{!xt}} templates, without having to examine the text very closely. That particular passage literally gives me a headache, and my vision is 20/15.)
• In the end, which manner of distinguishing correct examples from incorrect ones is judged "best" for MoS purposes is a matter of tastes, which are innately subjective. And of course, there's no arguing about tastes, as the cliché goes. (Just because it's a cliché doesn't mean it isn't true.) Notable exception: My tastes are impeccable, naturally, so they trump everyone else's, and I don't much care for the current templates.  ;-)

• Also, I don't understand at all what you're referring to by "a variation of approving or rejecting additions," so I can't comment on it. --Jackftwist (talk) 18:03, 28 July 2011 (UTC)

Sameboat: I like your initial suggestion above, at least as a starting point. [You posted your subsequent comment while I was composing this reply.] But placing the "Right" and "Wrong" after the example seems illogical or counterintuitive (for want of a better term). Putting those labels before the corresponding example seems pedagogically more sound because it signals the reader from the beginning what characteristic he (she) needs to be alert for that distinguishes the correct example from the incorrect one. I readily admit, though, that I don't know enough about wiki code/wiki markup/HTML/whatever all that gobbledygook is called to understand what your rationale means about "capitalization when making it into a template." Coming from you, though, I tend to think it's worth considering, but I should hope that problem could be overcome, even if the coding were more tedious. (Easy for me to say, since I won't be the one doing the coding!) Overall, though, this is a noteworthy suggestion, and I hope it gets the consideration it merits. We can quibble ad infinitum about which combination of background colors, fonts and colors, special symbols, etc., but this example is a solid point of departure.

• I'm not sure, though, how your Big Green Checkmark and Big Red X would be implemented with in-line examples. Could you show an example of how you see that working? --Jackftwist (talk) 18:03, 28 July 2011 (UTC)

Jeff: I completely agree that we don't need to apply any special template to examples shown in columnar form; they are indeed obvious enough. In-line examples are our main concern, although for consistency, IMO the same distinguishing template (whatever it is) should also be applied to cases where the right and wrong examples are shown one under the other. Other than columnar form, that format is by far the most unambiguous and easiest to see, even at first glance, and I hope it will continue to be by far the dominant format used in this section. Ideally, in-line examples should be used only when they are very short. (Yet again, see the "Hyphens" section for examples of flagrant abuse of in-line examples. I hate to keep picking on that section, but it's a formatting abomination.)

• But unlike you, Jeff, in this particular context, I like "anything that disrupts the normal flow." Rather than being distracting, IMO it's a distinct advantage to call attention to the contrasting examples as much as possible. For example, note the background color (and drastically different font) that seems to be the WP standard for setting off lines of code from the surrounding text, as in Sameboat's examples above; LaTex also does something similar by default when it renders its characters on the WP page. And the more I look at curly braces, the less I like them, despite my initial comments way above. (Cf. supra re tastes.) --Jackftwist (talk) 18:03, 28 July 2011 (UTC)

I must add that while MOS is generally accepted throughout WP, it's meant for encyclopedic article. Guideline or policy under other namespaces can look like a manual and be visually obtrusive, in a good way. The capitalization issue is that no wiki parser is able to check if the preceding context ended with a period, so we don't know if we should capitalize the right/wrong, we can just add another parameter for editor to change or just leave it all lowercases, though. The tick and cross images are the most straightforward possible, you just place it before the example, and the alt attribute of the image embedding syntax will display the replacement text (right/wrong) when image is disabled in reader's browser. The caption/title attribute functions as the mouse-over popup tooltip to indicate the meaning of the symbol. -- Sameboat - 同舟 (talk) 01:07, 29 July 2011 (UTC)

I'm afraid I must confess that I don't know anything about wiki parsers or alt attributes, so I don't understand most of what your comment means. (I'm a content editor, not a wiki markup, HTML, or whatever programmer.) And what do the capitalization issue and other namespaces have to do with this discussion? What does "add another parameter for [an?] editor to change" mean? (Are you referring to a human editor or to some kind of parser or bot?)

As for using tick and cross images, I don't have any conceptual objection to them per se, but would it be possible to use a (much) smaller image size that is more consistent with the font size of the text? The relatively large size of the images above really would be garish (to use JeffConrad's adjective), even overwhelming and distracting, especially on a page with a large number of them. (And we do have quite a few pages where a large number would be needed; e.g., see the long list of examples in Noetica's proposed revision to the "Dashes" section.) Also, wouldn't the large size look odd when used with in-line examples, especially if more than 1 pair of right-wrong examples appeared on the same line? (Or should that be "right–wrong" or "right – wrong"?) --Jackftwist (talk) 18:00, 2 August 2011 (UTC)

Check my work: Dashes

I proposed a move at Talk:Charlotte–Genesee Lighthouse#Requested move under my understanding of what "conjunction" and "disjunction" mean when deciding between hyphens and dashes, but now I'm concerned that that understanding may be incorrect. Could one of the local "experts" check my reasoning there and advise? Powers ^T 15:40, 31 July 2011 (UTC)

It's clearly a case for an en dash. I'm as confused as you are about what the "conjunction" means in WP:HYPHEN. I find no grammar or style books that use the term conjunction to describe the role of a hyphen. Maybe someone else knows of one that will clarify? Probably it just means a tight binding. Binding Charlotte tightly with Genesee would suggest an entity "Charlotte Genesee" being used as an attributive; the en dash would suggest that Charlotte and Genesee are both modifying lighthouse, in some relationship like "and" or "between"; in this case, indicating a location on the Genesee River, at or near Charlotte on land. Dicklyon (talk) 21:13, 31 July 2011 (UTC)

Okay, makes more sense now. Perhaps we ought to clarify "conjunction" and "disjunction" -- or remove them entirely -- from the policy page. Powers ^T 00:02, 1 August 2011 (UTC)

Yeah, we've gone round & round on that wording. 'Disjunction' is almost but not quite right. It's useful in the phrase disjunctive en dash, but as an actual explanation I think we need to spell it out: a union of independent elements, etc., not just "disjunction". — kwami (talk) 00:51, 1 August 2011 (UTC)

At least in the case of "disjunction", there are a few guides (by Garner) that talk about one usage of the en dash, of the sort where a slash (virgule) is sometimes used (see this), like "on–off switch" or "forward–reverse control", where it's one or the other but not both at once. But I don't see how "disjunction" is a good term for the more general case; where the relation is "and" or "between", I'd say it's more of a conjunction (a tying together, loosely). The using "conjunction" to describe the hyphen's role, we draw the line in the wrong place; time for a better description? Dicklyon (talk) 03:35, 1 August 2011 (UTC)

The best description is to be silent, and above all avoid made-up jargon. Any phrasing that inspires us to ask "what does that mean?" doesn't serve anybody. Septentrionalis PMAnderson 21:40, 1 August 2011 (UTC)

As far as I can tell, this relates to one sentence in the discussion of image filenames: A hyphen is used only to mark conjunction, not disjunction (for which en dashes are used: see below). Set aside the unsourced claim of fact here; even if it were true, why need it be said there? Our rule is that image filenames don't use dashes at all, because readers don't see them and editors (who do see them) want to be able to type them easily. Septentrionalis PMAnderson 21:50, 1 August 2011 (UTC)

There are other places to discuss your objections to the very idea, PMA; my question was about applying what's been agreed to for the time being. Powers ^T 15:38, 2 August 2011 (UTC)

Song vs. Tune

I should like to move that a guideline on the use of these terms be added to the MoS.

Americans seem to be the only English-speakers that refer to a piece instrumental music as a song; to everyone else it is a tune, or a piece (I know there are exceptions, like Mendelssohn’s Songs Without Words). Being English, I find it very confusing when reviews on amazon.com refer to "the songs on this abums", and I find later that the entire thing is instrumental. This is particularly so in idioms that could be either, such as Irish or Scottish folk music, or Flamenco.

Tune, on the other hand, is presumably comprehensible to everybody.

The guidelines say that "common ground should be sought where possible, words that would be confusing in other dialects should be avoided, etc."

So I should like to recommend respectfully that the word Song be avoided when referring to an instrumental piece. Paul Magnussen (talk) 16:13, 1 August 2011 (UTC)

I tend to think "melody, part of a song" when I hear "tune," (I'm from the U.S.) so perhaps it isn't as incomprehensible as you think. EDIT: I checked dictionary.com and it seems my impression is pretty close. "Tune" is not an ideal substitute for "song" or for "instrumental piece of music without singing."

That doesn't mean you don't have a point. MOS:COMMONALITY encourages using cross-English terms wherever possible (avoiding "bonnet" for "trunk of a car," using "fixed-wing aircraft" instead of "airplane" or "aeroplane," etc.) but we don't have a list of terms to be avoided in the MoS itself. It seems that your next step would be to create an entry for "song" in the article American and British English differences. Make sure that you have reliable sources that you can cite. For example, in my part of the U.S. the term "pigtails" is only used to mean bunches and "braids" means what others call pigtails. However, because this isn't 1. the dictionary definition or 2. a clear AmE/BrE division—lots of American sources, speakers and writers use "pigtails" the dictionary way—this meaning of "pigtails" doesn't merit an entry in this article. Darkfrog24 (talk) 17:43, 1 August 2011 (UTC)

The OED goes on at length with various definitions. I've always taken 2a as the main meaning, refering to a composition that sets a lyrical poem to a melody, as for Twinkle, Twinkle, Little Star. Different harmonies do not make the song a different song. Different singers do not make it a different song. Different lyrics do, as for The Alphabet Song which is sung to the same melody. Different recordings do not make the song a different song. Let us avoid conflating "song" with "performance", "recording", "track", "tune", "lyric" or anything else. The ideas are quite distinct, even if their vernacular usage is sometimes sloppy. LeadSongDog come howl! 19:16, 1 August 2011 (UTC)

So what is the word for a single popular-music composition of an independent melody, optionally with harmony or full backing accompaniment? "Composition" is too broad; "recording" doesn't apply to live performances, while "performance" doesn't apply to sheet music; "tune" doesn't incorporate the entire composition. Powers ^T 20:31, 1 August 2011 (UTC)

What about "piece"? It should be unambiguous in context; or, failing that, "instrumental piece" or just "instrumental"? For example, "Wipe Out" was specified as a Song originally. I changed it to Instrumental in 2009, and no one seems to have complained or reverted it yet. Paul Magnussen (talk) 20:43, 1 August 2011 (UTC)

That's fine when writing of a single work, but what about a collection of such works? Shall we say that an album comprised ten songs and one instrumental? Well, I suppose we could call them tracks, but then does it make sense to say "He performed several tracks from his album"? Or what if it's not from an album: "We heard several pieces from other artists" -- does that mean bits and pieces, or songs? Powers ^T 22:56, 1 August 2011 (UTC)

News to me that Felix Mendelssohn was an American; see Songs without Words. ;-> Septentrionalis PMAnderson 21:19, 1 August 2011 (UTC)

Septentrionalis, read the original comment more carefully! It depends what you are referring to. A folk musician will know various "tunes", which they will play with other musicians in various variations depending on the group. On a recording there are "tracks" (which used literally to be tracks in the vinyl days). A "piece" is normally a through-composed work, rather than a tune which can be adapted in performance. Surely it depends on the context whether "tune", "theme", "air", "song", "piece" etc are appropriate. Of course musicians - both popular and classical - often use technical terms in deliberately idiosyncratic ways. No-one thinks Hungarian Rhapsodies and Bohemian Rhapsodies are the same musical form. Paul B (talk) 21:26, 1 August 2011 (UTC)

And nobody in their right minds thinks "song" is the name of a musical form at all. This entire section is an example of why the Manual of Style is an attractive nuisance; every pedant thinks they can use it to make every editor Do It My Way, when in fact sane editors ignore this page altogether. Septentrionalis PMAnderson 21:36, 1 August 2011 (UTC)

Split hairs any way you want: the fact remains that using Song for something that isn‘t sung is confusing to many people, and should therefore be deprecated. Paul Magnussen (talk) 22:24, 1 August 2011 (UTC)

The difference between "effect" and "affect" is confusing to many people, too, but that doesn't mean we avoid the words. Powers ^T 22:56, 1 August 2011 (UTC)

Those words don’t mean different things to Americans and Britons. Paul Magnussen (talk) 06:09, 2 August 2011 (UTC)

"I've always taken 2a as the main meaning" leaves me thinking that there might be some peeves and personal preferences involved rather than a bona-fide case of people using the word wrong. Paul M and Lead Song Dog would be right to boldly edit articles so that they use more precise wording, but we should not work a rule banning this use of "song" into the MoS.

Of course the MOS doesn't ban the use of anything, it's a guideline. But we have scattered guidance on songs over many pages: Wikipedia:Manual_of_Style_(music), Wikipedia:Manual_of_Style_(titles), Wikipedia:Manual_of_Style_(lists_of_works), Wikipedia:Naming_conventions_(music) all pertain. The inactive Wikipedia:WikiProject_Music_terminology page also has relevant content. Of course, they are not entirely consistent. The "song" entry at the massive Virginia Tech Multimedia Music Dictionary is rather elegant in its simplicity - we might learn from it. LeadSongDog come howl! 15:45, 2 August 2011 (UTC)

While popular usage of the term "song" has possibly evolved, the lengthy discussion found in Grove's A Dictionary of music and musicians (A.D. 1450-1889) Vol. 3, Macmillan (1890) pp.584-632 starts out very simply. "In relation to the study oof music, a Song may be defined as a short musical composition, whose meaning is conveyed by the combined force of words and melody, and intended to be sung with or without accompaniment. The Song belongs, therefore, equally to poetry and music." Seems pretty clear to me. LeadSongDog come howl! 16:13, 2 August 2011 (UTC)

Again, if you guys can find sources showing that Americans use "song" one way and Britons use it another way, then put an entry into the article on American and British English differences—if this is a real difference, then you'd be doing everyone a favor—but that's as close to the MoS as this should get. Darkfrog24 (talk) 00:07, 2 August 2011 (UTC)

I think there already is (was?) a guideline suggesting not to call instrumentals “songs” somewhere in a subpage. Anyway, sometimes it can be tricky to decide whether it applies to a particular piece: is "To Live Is to Die" an instrumental? What about "The Great Gig in the Sky"? I say the former is and the latter isn't, but I can see why someone might disagree. ― A. di M.​_plé​^dréachtaí 15:31, 2 August 2011 (UTC)

More disruptive editing

I've reverted PMA a few times already, as he seems intent on editing the en dash section without seeking consensus first. We worked on this for months; the votes and discussion are on record, as are examples of a process that attempts to fine-tune it. Just inserting one's own view is not helpful. Will we be locking down MOS again? Or is there a simpler solution? Dicklyon (talk) 23:09, 1 August 2011 (UTC)

More of the usual. Let's have a few mere facts:

• We had a poll on this, at WT:Manual of Style/dash drafting, section 6b. It divided more or less evenly, between those for spaces around en dashes in dates, those against them, and those who were against such spaces in general, but hesitantly willing to consider them in dates.
• Of these objectors, Jeff Conrad had a whole sheaf of style manuals which opposed.
• Dicklyon started a conversation on this, here. Characteristically, most of you weren't invited.
• This resulted in what even Dicklyon admits is an impasse.
• Where there is no consensus, we should not make what we do not agree upon into a rule.
• ANevertheless, Dicklyon, as Owmer of this page, has rever-warred to imnpose an edit which does not refelct consensus, or sources..Septentrionalis PMAnderson 23:55, 1 August 2011 (UTC)

Yes Dicklyon, in what can only be described as a brain snap, Pmanderson has taken it upon himself today to edit the MOS—to let us all know how it really should be. I'm tired of being told how the MOS should be from someone whose self-declared motivation is the abolition of the document (e.g. continually trying to declare the MOS as historic—something the community has overwhelmingly rejected). Thank you Dicklyon for taking the time and trouble to try and improve the MOS for all, and I certainly felt invited when you wrote "Please join if you care". GFHandel ♬  00:21, 2 August 2011 (UTC)

And rhe other liar is heard from. My "self-declared" goal is the improvement of MOIS; if I wanted to abolish it, I would do what most Wikiepedians do and ignore it. But since Diclyon;s text is unsupported by consensus, style guides,,or usage, there will be az few who need mo other reason to approve of it. Septentrionalis PMAnderson 00:29, 2 August 2011 (UTC)

• Jeez, guys, tone it down. Don't refer to other editors as "liars". Don't accuse others of "owning" the page. Just discuss stuff before adding it to the MOS. If you want to add a statement that there is no consensus for a certain issue, then I think there should be a proposal made that that sort of statement be added to the text. Good Ol’factory ^(talk) 00:48, 2 August 2011 (UTC)

• PMA, the discussion on en dashes has moved here to Wikipedia talk:Manual of Style/endash spacing. You may discuss things there. The decision was to temporarily back out my effort at peace-making and use Dicklyon’s effort at consensus building as a placeholder while discussion is ongoing.

  Your edit (∆ edit, here) seemed well intentioned but, as you found out, just created conflict. Unfortunately, your response to being reverted was a childish editwarring comment written right into MOS, where you wrote The en dash in numerical ranges is unspaced; there is no consensus on times… (∆ edit, here).

  The world will not end if you have to stare at a placeholder effort as the issue is further discussed on the subpage to which I provided a link. However, were you to be allowed to persist at more of the same childish editwarring, MOS would get locked down again. The need for a lock-down is a clear sign that Wikipedia’s ability to reel in disruptive editors has broken down. I’m not going to wait around for others to try to set you up for a 3RR violation to prove that you are editing against consensus. You have no entitlement to a full three tries at editing against consensus. If you make any more waves on this issue as others go peaceably try to find the best solution, I’ll personally take you to ANI. Yes, that is a threat, promise, pledge, whatever you want to call it; it’s a fact. Greg L (talk) 00:55, 2 August 2011 (UTC)

• Thank you Good Olfactory. For the record, here's an example of Pmanderson's true motivations regarding MOS pages: At Wikipedia:Requests_for_arbitration/Date_delinking/Workshop#Cutting_the_Gordian_knot Pmanderson propsed: "WP:MOSNUM is declared historic. It shall be tagged {{historic}} and kept protected;" and "Let's get rid of it". (BTW, that proposal received no support in the subsequent discussion.) I would appreciate if the (incorrect) allegation of "liar" is retracted, and if not, could an admin please step in and address the injustice. GFHandel ♬  01:06, 2 August 2011 (UTC)
  • The word "liar" should certainly be retracted; a much smaller concern is that "declare the MOS as historic" doesn't quite match "WP:MOSNUM is declared historic", and MOS should be changed to MOSNUM. Art LaPella (talk) 01:26, 2 August 2011 (UTC)
    • Sure, let’s deal with that little jewel head-on. On the subject of personal attacks, PMA: your And rhe other liar is heard from seems to have gotten the goat of GFHandel. I tend to let stuff straight the heck off the 2nd-grade playground role off my back like water off a duck. Wikipedia being the all-volunteer joint it is tends to have a spectacular degree of inconsistency in its enforcement on that sort of thing. To a large extent, I think the community gives wide latitude to characters who are seen as generally incapable of conduct expected. All I can say is you didn’t do yourself any favors with that personal attack on GFHandel. It’s becoming increasingly clear to me why your presence at MOS is closely associated with lockdowns. I’m tempted to start the process of dealing with you using a broad brush; it’s not going to take much. I suggest you either get an attitude transplant real pronto here, or take a wiki-break. Greg L (talk) 01:32, 2 August 2011 (UTC)
    • Art, you would be right if Pmanderson's real aim was only to declare MOSNUM historic (while leaving all the other MOS pages standing)—which all those who have had long-term dealings with him know isn't correct. No, no, the quote I gave is more than indicative of his real motivations around these parts. The more I think about his statement "if I wanted to abolish it, I would do what most Wikiepedians do and ignore it" the more I find it utterly agreeable, and I support him following that strategy to attempt to abolish the MOS. GFHandel ♬  01:43, 2 August 2011 (UTC)
      • Who gives a crap what PMA thinks in that regard? It’s way outside of the mainstream so it’s a waste spending time to contemplate its Great Meaning.®™© It’s time to move on and discuss things civilly at Wikipedia talk:Manual of Style/endash spacing. If PMA behaves himself there, fine. If not, fine; I know how I plan on dealing with that too. Greg L (talk) 02:22, 2 August 2011 (UTC)
        • I don't. If you didn't look so much like mirror images, this would have probably been resolved years ago. Art LaPella (talk) 02:42, 2 August 2011 (UTC)
          • Who are you talking to and what are you alluding to, Art? Greg L (talk) 03:56, 2 August 2011 (UTC)
            • "I don't" refers to "I know how I plan on dealing with that too." As for "if you didn't look so much like mirror images", well ... let's just leave that at "you plural". Art LaPella (talk) 04:12, 2 August 2011 (UTC)
              • Well, “Awesome Wikipedian” or not (I just looked at your user page), that just comes across as your galloping in here on your whiter-than-white tall steed, it’s nostrils flaring in the morning mist, and having your steed rear up while you pronounce nebulous smack-downs on the little minions fighting in the trenches. It’s transparent and doesn’t impress. If you have something constructive to say, say what you mean. Otherwise, spare us the high-road posturing; M’kay? Greg L (talk) 04:21, 2 August 2011 (UTC)
                • Aw, go fight on in the trenches. Art LaPella (talk) 05:27, 2 August 2011 (UTC)

As an update, I have blocked PMAnderson for a week. Trying to undermine and weaken the dash consensus (like this and this) that has required months of discussion is just not on. Plus subsequent edits show a lack of interest in discussing this constructively. Casliber (talk · contribs) 09:51, 2 August 2011 (UTC)

Thank you for watching. Nobody else is. I bet they still can't agree on dash spacing. Art LaPella (talk) 14:43, 2 August 2011 (UTC)

• Yes, thanks. I hope other admins don’t try to second-guess you and pull the rug out from under your block, Casliber. Unless the other admins are going to invest the time to understand the history and pattern of editing and temperament of PMA, they’re not going to understand why WP:MOS actually had to be locked down for days-on-end; lock-downs are a clear sign that something was way-wrong. Greg L (talk) 15:24, 2 August 2011 (UTC)

Links in section headings

Headings should not normally contain links, especially where only part of a heading is linked.

Just out of curiosity, can somebody explain the reasoning behind this policy? –CWenger (^ • @) 22:19, 2 August 2011 (UTC)

Big horrible blue links make it look ugly. There may be technical reasons but for me this is the main thing. violet/riga (t) 22:29, 2 August 2011 (UTC)

Search Help:Section for "internal link in a section heading". Any other reasons? Art LaPella (talk) 22:31, 2 August 2011 (UTC)

And their effect on an article is ugly. Dicklyon (talk) 22:33, 2 August 2011 (UTC)

Original titles for works in foreign languages

Hello The issue presented here started at a discussion for the naming of the Caso cerrado article, questioning whether English orthodox was to be applied to a Spanish-language title with thus different orthodox rules. After much heated discussion we came to the conclusion that Wikipedia policies do not seem to have specific guidelines when it comes to the proper formatting for titles in foreign languages. Two Wikipedia policies were brought up to support the use of English orthodox in any title, regardless of language, but I chose to disregard them based on the following:

• USEENGLISH was used as a policy to support the move from "Caso cerrado" to "Caso Cerrado." However, this policy states that if a published work has an official English title in addition to a foreign language title, the English title is to be used. For example, the policy suggests that "Pan's Labyrinth" is to be used instead of "El laberinto del fauno" because the former is an official English title. It does not, however, suggest that English orthodox is to be applied to a foreign-language title with no English alternative.
• COMMONNAME was also presented, stating that common names are to be listed as they're more commonly formatted in English-language sources. However, foreign language titles do not qualify as "common names." The policy refers to common name terms such as "personal computer," "apple tree" and "beach towel," not titles of published works.

Wikipedia houses thousands of articles about foreign language titles in Spanish, French, Finnish, Italian, Galician, etc. About half of these articles apply English orthodox to foreign language titles, and the other half apply the orthodox associated with the language the titles are in. It is crucial that specific guidelines are made available, so users can determine which format to use. Generally speaking, English orthodox is not to be applied to foreign languages in the same manner that Spanish/Italian/French orthodox is not applied to English language terms. Regardless of this being the English Wikipedia, when titles or quotations in foreign languages are presented, they're still in a foreign language, and that language's orthodox should be respected, if nothing more, for common courtesy.

The Internet Movie Database is often disregarded as an unreliable source because of its high number of user-generated data. However, their very specific and strict guidelines are written by their knowledgeable staff, and they claim "Capitalization [of titles] is language dependent." (IMDb: Guide - Capitalization and character sets) In other words, regardless of IMDb being an English-language site, titles are to be presented and formatted according to the orthodox accepted by their respectful languages. Many other sites adapt these same principles and are open to them when suggested (such as in title corrections, etc), even if not always written as part of their guidelines.

The bottom line is that there are currently no specific guidelines regarding proper formatting for foreign titles, this leads to confusion and in cases, pretty heated discussions with no evidence to back up either side.

Personally, I think the same rules should be applied here than in other Wikipedias: A foreign language's orthodox should be respected whenever and wherever that language is used. No foreign language becomes English simply because it is presented as part of a source with English as its dominant language.

Many thanks. T.W. (talk) 20:43, 27 July 2011 (UTC)

This isn't really an AN issue; WP:VPP and/or WT:MOS would seem to be better locations for this. I have an opinion on it, but raising it in a better forum is more likely to generate a meaningful response. The Blade of the Northern Lights (話して下さい) 23:05, 27 July 2011 (UTC)

My own take on the matter would be to use Pan's Labyrinth but make sure that El laaberinto del fauno be used in the lead. I would not, however, invent an English-language name for Wikipedia. Darkfrog24 (talk) 01:16, 28 July 2011 (UTC)

Thanks for your imput. Remember, however that the issue here is whether or not English orthodox is to be imposed on foreign language titles. For example, Spanish, Italian and French languages capitalize the first word and any proper nouns within a title ("El laberinto del fauno"), whereas English-language orthodox is to capitalize the first word and any keywords, excluding function words ("El Laberinto del Fauno"), making said titles incorrect in their respective languages. My argument is that English language orthodox does not and should not apply to foreign language titles, so "El laberinto del fauno" is correct. Regardless of being presented in the English Wikipedia, "El laberinto del fauno" is still Spanish and should abide by its rules, not English ones. T.W. (talk) 02:09, 28 July 2011 (UTC)

Just as a note, I think the word you're looking for is orthography, which is about writing, not orthodox, which is usually religious in connotation (unless someone's talking about right-handed boxers).

As for the question, if a foreign-language title of the work is being used as an article title because no commonly used English version exists (and making up one, even a direct translation, is original research), then the orthography of the original language should be used, as it accurately represents the original title, presuming its compatible with the Latin alphabet. (Transliteration of languages written in other alphabets is tougher, but that's a case of following the sources; we shouldn't make up transliterations, either.) oknazevad (talk) 05:53, 28 July 2011 (UTC)

I have to disagree. Translation is not the same as original research; in fact it can result in copyright being breached if the rules aren't followed.

We also need to avoid creating "one rule that fits all". Names of "works" are one thing but other titles should be appropriately translated and, yes, English capitalisation or lack of hyphenation used. "Theodor Heuss Bridge" is a perfectly acceptable name for the German "Theodor-Heuss-Brücke" even if it were an obscure name never before mentioned in English. This is standard translation practice and we will produce reams of unintelligible titles if we do otherwise. Also there is a need for consistency. If famous mountain passes are converted from "Foopass" to "Foo Pass" it would be inconsistent and illogical not to do the same for a pass never before mentioned in English books. That creates a 2-tier system. To sum up, a blanket rule here is dangerous, which is why WikiProjects for countries have developed naming conventions to guide editors. --Bermicourt (talk) 06:14, 28 July 2011 (UTC)

(edit conflict)

FWIW, "orthodox" can be, and often is, used in a non-religious context as "adhering to what is commonly accepted, customary, or traditional". And IIUC "orthography" of English, or any language, can be seen as codifying the orthodox rules of writing.

That aside. I would agree that, for works that do not have English titles, the original titles be used and their use be consistent with the rules of grammar of the original language, up to an including diacritics. The only caveat would be to find the "romanized" version for languages not using a Latin character base.

@Bermicourt: If the Wikipedia editor is translating it's going to look like OR. If they are citing an English language source that uses the translation, it's sourced. Also, based on the original post, we are dealing with titles in relations to "works" - the first is a Spanish language TV show and the second is a film originaly released in Spanish. I'm not sure a premptive "This should not cover locations" is needed at this point.

- J Greb (talk) 06:30, 28 July 2011 (UTC)

I should mention that this would apply to latin alphabet-based languages (Portuguese, Romanian, Spanish, Italian, German, Galician, etc.), non-latin alphabet languages such as Chinese, Japanese and Arabic would naturally, if using their native titles, be romanized as opposed to using their characters (for article titles, the original non-latin title would appear in the introduction). Remember this is only for titles of published works, meaning television shows, programs, books, songs, albums, movies, etc. T.W. (talk) 16:59, 28 July 2011 (UTC)

This is at least the third forum in which this has come up; I met it at WP:AN.

My view is that we should do what other English-speakers do. This article has three sources in English (one of them a Hispanic News reporter, who presumably knows and cares about Spanish conventions). All of them use the English convention for marking a proper noun: Casa Cerrado. We should not make up conventions; they will not be understood, which is our goal here. Septentrionalis PMAnderson 17:52, 28 July 2011 (UTC)

My view is the same as PMAnderson's. Just follow what the majority of English-language sources do when they are discussing the subject. Jenks24 (talk) 18:05, 28 July 2011 (UTC)

Nothing is being made up. Foreign languages have specific regulations for capitalization of titles in their respective languages. Spanish rules go as follows:

"Capitalization in Spanish is sparse compared to English. In general, only personal and place names, some abbreviations (e.g. Sr. López, but señor López), the first word (only) in the title of a book, movie, song, etc., and the first word in a sentence are capitalized, as are names of companies, government bodies, etc. Names of nationalities or languages are not capitalized, nor (in standard style) are days of the week and months of the year." from Spanish orthography#Capitalization

It is more than safe to assume that English-language authors of Internet articles are less-than-familiar with foreign language orthodox, and not reliable sources when it comes to proper formatting. There is no dispute here regarding which form is proper: "Caso cerrado" abides to proper Spanish orthography, "Caso Cerrado" is incorrect. "Caso cerrado" is not English, it's Spanish, and should follow the latter's grammar rules. The same goes for every other latin-alphabet based languages.

The Internet Movie Database may not be a reliable source when it comes to their user-generated sections (trivia, goofs, etc.), but they DO have strict policies regarding capitalization in all languages according to said languages' proper orthodox, which none of the sources listed do. Based on those policies, they list the title correctly as "Caso cerrado." [5] T.W. (talk) 19:24, 28 July 2011 (UTC)

Major problems with relying on English-language sources when choosing which orthodox to follow:

• The vast majority of sources will not have neither specific nor strict guidelines relating to proper orthodox use. Instead, they'll be formatted according to the specific author's personal preference instead of any set guidelines. This makes those sources unreliable when it comes to choosing which is correct.
• Hundreds of foreign-language titles will have zero English sources (aside from the standard IMDb which does have orthodox guidelines, not chosen by specific authors or users), leaving countless article titles vulnerable to further indecission and conflict. Also, in many cases, English sources are bound to contradict one another, once again, leaving room for indecission. For these reasons I think it is crucial that a decission be made between applying English orthodox, making titles incorrectly formatted in their respective languages, or abiding to their proper orthodox, as accepted by their respective language. T.W. (talk) 21:01, 28 July 2011 (UTC)

After some research I discovered that there are in fact somewhat flaky guidelines for adapting foreign capitalization into foreign language expressions: Capitalization of expressions borrowed from other languages. "Loan words" such as the French-adapted English title for "Les Miserables," naturally follow English conventions, but foreign words in the original foreign titles should follow the conventions associated with that language. For example, an English version of "Caso cerrado" does not exist in any form (subtitles do not qualify as 'English' version, assuming they even exist), therefore the title has neither been adapted as "Caso Cerrado" into English, nor translated from Spanish as "Case Closed." It remains a Spanish-language title for a Spanish-language television program, of which no English version exists. T.W. (talk) 20:43, 29 July 2011 (UTC)

That looks like a sensible guideline. Is anyone against it? Dicklyon (talk) 23:49, 29 July 2011 (UTC)

And did we agree that we're talking about orthography, not orthodox? Dicklyon (talk) 23:51, 29 July 2011 (UTC)

"Les Miserables" is a title adapted into English, it has become part of the English language and is thus an English title, it also does not apply. T.W. (talk) 19:42, 29 July 2011 (UTC)

And how do we tell the difference? Ask the infallible intuition of Taran Wanderer? Les Miserables has existed a century or so longer than any television series, and is more frequently mentioned in English; but "Casa Cerrado" is equally adopted. Septentrionalis PMAnderson 20:13, 29 July 2011 (UTC)

"Les Miserables" has been published in English with its original French title, thus the French title has been adapted into English. Caso cerrado has not been adapted into English in any shape or form, it remains the title to a Spanish-language television show. There's no English version with any title, neither adapted nor translated. THAT's how the difference is made crystal clear, and yes, you can trust Taran Wanderer, though I won't make myself a Wikipedia policy. T.W. (talk) 20:22, 29 July 2011 (UTC)

Les Misérables is the capitalization used in French as well as in English. john k (talk) 02:32, 4 August 2011 (UTC)

In which case Victor Hugo likely meant "Misérables" to perform as a proper noun, as opposed to the common "misérables," as was adapted into Spanish (miserables) and Italian (miserabili). Very often, titles on the covers of these works will adapt the 'capitalize all keywords' rule, but only for aesthetic reasons, their official titles will still follow the language's ruling, regardless. This is common in movie posters and DVD covers, etc. T.W. (talk) 04:01, 4 August 2011 (UTC)

Poll

Let's continue the discussion above, but let's organize each contributor's decision below as a poll to get a clear picture of where we stand. Those in

...favor of applying English conventions to titles of published works in foreign languages

(e.g. German: "Märchen von Einem, der Auszog das Fürchten zu Lernen" as opposed to "Märchen von einem, der auszog das Fürchten zu lernen" or Italian: "Un Medicco in Famiglia" as opposed to "Un medico in famiglia")

...favor of abiding to foreign language conventions in titles of published works in foreign languages

(e.g. German, first word and all nouns capitalized only: "Märchen von einem, der auszog das Fürchten zu lernen" or Italian, first word and proper nouns capitalized only: "Un medico in famiglia")

1. Woodstone (talk) 06:49, 29 July 2011 (UTC), with the remark that this would follow the best sources for those titles as well
2. User:Martinvl Martinvl (talk) 07:29, 29 July 2011 (UTC)
3. This is, in fact, what reliable sources do. If we're going to have a manual of style, it is kind of pointless to just have it say "follow reliable sources." The point of it should be to do some research into how reliable sources deal with certain frequently arising issues and then recommend that we do that. This is a clear instance of that - I have never seen a bibliographical entry in a scholarly work which does not follow the title capitalization conventions of the language the cited work was originally written in. john k (talk) 17:52, 29 July 2011 (UTC)

   An argument worth answering: we are not a bibliography, and Cosí Fan Tutti is normally so capped in running prose. But this is the second comment which does not actually disagree with the third position below; an argument that we need not consult reliable sources only works for edits that aren't based on reliable sources. Septentrionalis PMAnderson 17:58, 29 July 2011 (UTC)

   I don't necessarily disagree with the position below, but I do think it makes sense to be somewhat more systematic than to just look up reliable sources for each individual case. As to Cosí fan tutte, a google news search suggests that it is frequently not capped in running prose in the manner you prefer, although it sometimes is. john k (talk) 00:11, 30 July 2011 (UTC)

   I agree; by adapting this policy, specific cases would no longer be a problem. All languages (though the conventions page seems to be unaware of this) have specific guidelines for capitalization that we can turn to for any title. In the case of 'loan' titles ("Les Miserables," "Cosi Fan Tutti" and "La Isla Bonita"), there's no question we should use English-language guidelines. Adapting this policy would make the lack of sources not be a problem. What's clear is that we're all against the first option, that of forcing English conventions on foreign titles, right? T.W. (talk) 01:16, 30 July 2011 (UTC)

   Les Misérables has that capitalization in French as well as in English. As to Cosi fan tutte, I'm not sure how there's "no question" when, as I pointed out, many news stories use the Italian capitalization. john k (talk) 02:30, 4 August 2011 (UTC)

4. T.W. (talk) 19:22, 29 July 2011 (UTC)
5. Yes, when there's no obvious English version to choose. Dicklyon (talk) 23:57, 29 July 2011 (UTC)
6. Though there may wind up being a case by case discussion as to where a work such as a play, film, book, or similar has had an English language edition published with an anglicized title. It is also important to remember that this is not restricted to article titles, but also covers how the titles of works are presented within the whol of articles. Citing WP:CAPS and WP:COMMONNAME is all well and good when trying to figure out an articles title. But those hold very little weight with the style used with in the article. - J Greb (talk) 00:14, 30 July 2011 (UTC)
7. In a very few exceptional cases there will be an established name used in English-speaking countries to refer to a foreign-language version of a work (a name is established by the consensus of the language community, not a few dozen Google hits). Such well-known titles will often merit an entry in other reputable encyclopaedias and we can follow their example:

• the Luther Bible (no italics)
• Historiae animalium

In most cases there will probably be no established English name (differing from the foreign name), and foreign conventions, including capitalization, should normally be followed.

Normally, the reader should be able to assume that an English name using English spelling and capitalization refers to the English version; any foreign language version should use the foreign name; and the spelling and capitalization rules appropriate to the language should be used:

• Albert Einstein: schepper en rebel is the Dutch version of Albert Einstein: Creator and Rebel
• Der arme Heinrich
• Under der linden by Walther von der Vogelweide
• Der Hauptmann von Köpenick
• Die Täter sind unter uns

Unless editors are agreed that there is good reason to do otherwise, with a few exceptions such as all caps we should use the name exactly as it is printed on the front page (or in the national library catalogue). How The News of the World spells the works of Walther von der Vogelweide is of little interest. We should not attempt to apply English conventions to foreign words.--Boson (talk) 21:58, 2 August 2011 (UTC)

....favor neither position above, but do what our English-language sources do.

1. Septentrionalis PMAnderson 20:31, 28 July 2011 (UTC)
2. per WP:COMMONNAME. Blueboar (talk) 15:46, 29 July 2011 (UTC)
3. I don't see any reason not to follow WP:Naming conventions (capitalization)#Capitalization_of_expressions_borrowed_from_other_languages, which says, "If the article is about a work in a foreign language (such as a book or other written work, movie, album, or song), using the capitalization found in most English language reliable sources is recommended." WhatamIdoing (talk) 23:13, 29 July 2011 (UTC)

   Because only a portion of foreign-language titles have reliable English sources to begin with. Few, if any, follow any set guidelines for capitalization, and are up instead to the preference of the author, who in many cases will assume that English-language capitalization is universal. Thus not 'reliable' sources in this manner. It is much easier to just follow the languages' own standards, since reliable sources should be doing that to begin with. T.W. (talk) 23:29, 29 July 2011 (UTC)

   In such cases, there is always the question whether we should have an article on the subject to begin with; there are notability requirements for books and prograpms. Some will be notable, despite never being discussed in English; for that minority, we can sometimes proceed by analogy; if not, then thee is no extablished English style, and the style of the original language is the reasonable thing to follow. But we are discussing a minority of a minority. Septentrionalis PMAnderson 23:38, 29 July 2011 (UTC)

4. When English sources agree, we can use them; the problems are mostly only when there are no English sources for a title, or where there is no obviously accepted English version. So follow WP:Naming conventions (capitalization)#Capitalization_of_expressions_borrowed_from_other_languages when possible. Dicklyon (talk) 23:59, 29 July 2011 (UTC)
5. Wikipedia needs to follow the practices of other quality English-language publications and not pretend we can change the very alphabet used to write the English language.

   A trema ( ¨ ) is something I can type from my Mac keyboard (and just now did) precisely because Apple recognized that certain diacriticals are commonly used in English and endeavored to make it easier for Mac users to deal with common English-language practices. Wikipedia has zero business pretending it is *being enlightened* with stuff like “Nguyễn Ngọc Loan” (the Vietnamese street execution dude) since none of those diacriticals can reasonably be considered as part of the English language. It doesn’t matter what our motives are, or how worldly and wise we think we are as pure-volunteer wikipedians; our only business is to look to best practices of quality English-language publications and follow what they do.

   
   → Shooting from the hip a bit here, I’d say we couldn’t go too wrong if we adopted a “two out of three” proposition after looking at how National Geographic, The New York Times, and Encyclopedia Britannica handle these things.←

   Oh… and for those who might cite how technology like Unicode allows wikipedians to show the rest of the world how to do things smart-like, uhm… no; type foundries have long offered every diacritical known to man and the aforementioned publications could use them whenever they want. It’s not about *technology*.

   To PMA: Your name all by itself at the top of this doesn’t constitute a persuasive argument capable of swaying opinion; it’s a name on a list. Let’s see you use that brain of yours to advance some cool-beans reasoning that makes other editors arch their eyebrows, pout out their lower lip, and “Huh… he’s got a point.” That sort of thing factors into gauging the quality of each side’s position when discerning a consensus. Greg L (talk) 04:50, 30 July 2011 (UTC)

   
   P.S. BTW, I hope it was just an oversight in the lede of this proposal, when only German and Italian examples were mentioned and then the scope of the actual proposal itself was broadened to “foreign languages”. I note also the exceedingly magnanimous and wholesome wording of “abiding to foreign language conventions”. It still boils down to the fact that England, where our language came from, had its share of cultural influence from Germany, Italy, and France and this influence crept into common usage of the language used in England. (Then those darned Yanks went across the pond and changed some spellings, adopted other words for some common nouns, and likely evolved the adoption of certain diacriticals.)

   As has been shown a hundred times before, it is folly to think we can exploit Wikipedia in hopes of changing how the world works because we think there is a better way; particularly when we hope to change the very alphabet we use to write the English language. Were it me, I would counter with my own RfC and in the proposal, write Shall Wikipedia abide by common English-language conventions used by high-quality respected publications.

   Judging from the practices here by The New York Times and here at National Geographic, (two out of my above-mentioned three highly respected publications), Wikipedia would follow their lead and spell it François Mitterrand. And in the case of Nguyen Ngoc Loan, we’d probably find ourselves again following what National Geographic does as I wouldn’t be surprised if at least one of the remaining two sources I suggest we look towards (The New York Times and Encyclopedia Britannica) would handle his name the same way. Greg L (talk) 14:40, 30 July 2011 (UTC)

   If this approach is followed, then care must be taken to ensure that the "respectable sources" come from both sides of the Atlantic. Martinvl (talk) 20:26, 30 July 2011 (UTC)

   That works for me. Any country that has English as its primary language works for me. That would include England and that “barbie” country where even the moon is upside down. Note that Encyclopædia Britannica has strong roots in Scotland and is owned by a Swiss guy. We wouldn’t be exactly turning our backs on the old world by including it in the three I proposed—besides, it’s the most respected encyclopedia out there. And National Geographic is published in 33 languages worldwide. We can rely on them to know about languages in general, and seek guidance from them on English-language issues directed to a readership that considers itself to be worldly. Greg L (talk) 22:04, 30 July 2011 (UTC)

6. Per WhatamIdoing. The original capitalization should be fallen back upon when there's no established Anglicization. ― A. di M.​_plé​^dréachtaí 23:16, 30 July 2011 (UTC)

   There's only "stablished Anglicization" when the work in question is adapted into English and its original title is adopted. If "Caso cerrado" had been dubbed into English and its name had been borrowed, "Caso Cerrado" would become proper English. However, the fact that Variety.com (in some articles and not in others), and others use "Caso Cerrado" is not "Anglicization stablishment," just an assumption that English/Portuguese capitalization is universal. For an earlier argument about only German and Italian being used as examples, I did not think it necessary to list an example for every roman-alphabet-based language. Of course Spanish, Portuguese, Finnish, Romanian, Galician, Catalan, Sicilian, Bask, Estonian, Dutch, etc. are included in this proposal. T.W. (talk) 00:14, 31 July 2011 (UTC)

   I was talking about the rule in general, not about any particular case. I don't even know what the hell "Caso cerrado" is. ― A. di M.​_plé​^dréachtaí 00:29, 31 July 2011 (UTC)

   It's one of the examples we've been using. There's no English version of it, therefore it has not been 'adoted,' regardless of any English sources choosing to do so. I'm also speaking in general, continuing to use this case as example. There's confusion between titles that have been rightfully adapted into English (such as "Cinema Paradiso," or has it?) and others that have simply been formatted wrong in articles. T.W. (talk) 00:43, 31 July 2011 (UTC)

   Where follow the RSs doesn’t fully give us guidance, use common sense will generally fill in the gaps. Greg L (talk) 16:57, 31 July 2011 (UTC)

   Prioritizing common sense would be a good practice. Common sense tells me that when an article title (or title within the article) refers to a title in its original language, said language's conventions should be used (such as with "Caso cerrado" or "El laberinto del fauno"), when it refers to a title of a work adapted into English (whether translated: "Pan's Labyrinth" or borrowed: "La Isla Bonita"), then English conventions should be used. Simple as that. T.W. (talk) 17:46, 31 July 2011 (UTC)

   I think MOS only needs to cover the broad principal and the vast majority of common stuff. There are bound to be plenty of details such as “when an article title (or title within the article) refers to a title in its original language.” For the most part, looking towards precedence established by respected English-language publications should give wikipedians guidance as to best practices. In other cases, where guidance is elusive, editors can establish a local consensus without trying to resolve everything at WP:MOS.

   I have no problem with the examples you provided ("Caso cerrado" or "El laberinto del fauno"). If “said language conventions” of the foreign language means adoption of diacriticals not used by high quality English-langauge publications like The New York Times, National Geographic, and Encyclopædia Britannica, then no, I don’t agree with you whatsoever. It wouldn’t matter if there was a Broadway stage play on the Vietnam War titled “Nguyen Ngoc Loan and His 38-caliber Pistol,” and the play is named after a Vietnamese book titled “Nguyễn Ngọc Loan: Our ‘Dirty Harry’ ”, we can’t have mere wikipedians trying to change the English language and use the Vietnamese-language diacriticals.

   Even if these efforts are given wholesome slogans like “Peacefully abide by all the alphabet conventions of world without *hate language* and with love for everyone”, we can’t be running off doing our *own thing* and try to expand the English alphabet beyond what is used by other respected English-language publications. That’s beyond the scope of what mere volunteer wikipedians with way too much time on their hands can rightfully do. Greg L (talk) 22:05, 31 July 2011 (UTC)

   Well, I would hope such major and respected publications would have their foreign titles properly adapted, but in any case, I doubt any of them will have any sort of publication on most of these works, since very few of them have had a significant impact on English-speaking regions. I'm sure those publications you mention have specifically-set guidelines (like IMDb.com does) when treating foreign titles, which cannot be said for a vast majority of 'reliable' English sources, such as Variety.com, where the capitalization is based on the author's own limited knowledge/personal preferences, as opposed to any guidelines, thus making their articles an unreliable or flaky source when choosing which format is appropriate, specially if fair knowledge on the language and common sense can be applied to prove them wrong. T.W. (talk) 22:32, 31 July 2011 (UTC)

7. Hesperian 00:19, 1 August 2011 (UTC)

Spaces in "XXX – United States relations"

See Category:Bilateral relations of the United States. There are tons of articles like this with spaced en dashes, which was previously what MOS:DASH recommended; now it recommends unspaced. Are we really wanting to make this much of a change? Will we be moving all those now? Any strong opinions one way or the other? Should I go ahead and RM them, or should I extend the requested optionality of such things into a new en dash change proposal that I'm working on? Dicklyon (talk) 03:04, 29 July 2011 (UTC)

I would suggest just ignoring the issue. Apply the new guideline to new articles and let sleeping dogs be. Blueboar (talk) 15:50, 29 July 2011 (UTC)

This discussion of not making trivial title changes would seem to apply. Septentrionalis PMAnderson 17:05, 29 July 2011 (UTC)

Yes, this can be taken to be a particular case of that general question that got no answer. I agree with you that sometimes "article titles should be changed for small benefeits," as you said there. Ignoring the issue will lead to years of churn, I expect, so I'm seeking to see if we can agree on a direction to reconcile guidelines and practice, for a cleaner resolution. Dicklyon (talk) 20:29, 29 July 2011 (UTC)

• Just noting that the spacing of en dashes in bilateral relations articles and categories has recently become a source of confusion for quite a few at CfD and RM. It would be great if this could be clarified in MOS:DASH somehow. Jenks24 (talk) 15:11, 30 July 2011 (UTC)

• Comment. I don't really care which way is decided on this issue, but whoever is doing the consensus-building needs to decide which it is and then implement it for old articles as well as new. Otherwise confusion results because editors see the "old" articles which use the old system and assume it is still in force and so they create new articles using the old format. Editors are far more likely to pattern article names after existing article names rather than consulting WP:DASH. Average editors probably also have no sense of how "old" an article is without specifically checking, which they presumably have little reason to do. So please decide on what the convention is and then implement it. If you're not prepared to implement it on old articles, don't change the convention. (These issues are not merely theoretical but as noted have arisen at category rename discussions, since category names generally follow the name of the article of the same name.) Good Ol’factory ^(talk) 23:02, 1 August 2011 (UTC)

Exactly why this kind of change should be minimised. See the discussion on changing the spacing rules for month-day dates but not for month-day-year dates: fiddly and with huge reverse-engineering problems. Tony (talk) 01:47, 3 August 2011 (UTC)

But we did change it (unless you interpret this as being like a date?); so should we change it back? Or clarify what we meant? Dicklyon (talk) 01:50, 3 August 2011 (UTC)

To tell you the truth, as someone who wasn't involved in the discussion to revise, when I read what currently exists, I'm not really clear on whether the rule was changed or not. It used to be that there should be a space in "XXXX – United States relations". So what is the situation now? Space or no space? Or was that left ambiguous due to disagreements on this point? If there is a position, it would help to include some examples using country names. Good Ol’factory ^(talk) 02:11, 3 August 2011 (UTC)

I thought we had moved away from most spaced en dashes as a compromise. But if so, the compromise remains too unclear. Dicklyon (talk) 02:50, 3 August 2011 (UTC)

The reason I am confused is this sentence: "The en dash in a range is always unspaced, except for times and dates (or similar cases) when the components already include at least one space." Why isn't "XXXX – United States relations" a similar case when the components ("United States") already includes at least one space? I agree that it could be clearer. I don't mind working to "correct" some of the old spacing issues in the bilateral relations articles and categories so that they conform to the new standard, but it would probably be best if there was something clear that I could point to that justifies the change. Right now it's ambiguous, if not misleading. Good Ol’factory ^(talk) 03:02, 3 August 2011 (UTC)

In reviewing Noetica's draft and comments on it, he clearly intended the spacing only in ranges, not in other symmetric relations. See this diff where he says "note that the sense of the dash differs. In one it shows a range, in the other it shows a relationship between two elements. Given that difference in meaning, it is well that the dash should at least behave differently with respect to spacing. That is at least as arguable. In the end, I believe the position I have incorporated reflects the clear majority in the voting." This matches my understanding at the time. However, now I think we ought to consider also specifying spaces in other uses, such as in Mexico – United States relations. By the same logic, I suppose I interpret "or similar cases" as meaning ranges that include spaces, in general, including 28 mm – 17 m, though that was not what I was thinking at the time. If you go back to Kotniski's draft Wikipedia:Manual_of_Style/Dash_draft, you find much more explicit description of spacing in ranges: "The dash is usually unspaced, but if the elements being linked themselves contain spaces (as with full dates), then the dash is spaced." There was some support for this draft, and I don't find any specific opposition to the spacing. In the comments at Wikipedia_talk:Manual_of_Style/dash_drafting/discussion#A_draft, for example, PMAnderson objected a bit, but not to the spacing provisions at all. Kotniski's also had unspaced en dashes in other pairs, including "the New Zealand–South Africa grand final". So Noetica's and Kotniski's are essentially the same, and got no objection to the spacing provisions besides from Jeff (until after the fact, when PMA joined him to disrupt things). That doesn't mean we shouldn't try to fine tune it, but in my experience, that may be difficult. Dicklyon (talk) 04:29, 3 August 2011 (UTC)

Thank-you, that's a great summary of the background. I understand now. I definitely would agree that in non-range usages, it is unusual to space the dash, so I would have no objection to clarifying this issue about non-range spacing. Good Ol’factory ^(talk) 04:51, 3 August 2011 (UTC)

Well, it's not at all unusual in WP article titles! Dicklyon (talk) 05:05, 3 August 2011 (UTC)

Very true. The previous guideline was implemented quite well, which is a good sign, actually. It means established guidelines actually do get followed. Good Ol’factory ^(talk) 09:57, 3 August 2011 (UTC)

I think it should be explicitly stated in the guideline that these en dashes are unspaced even if the parts contain spaces (particularly since this represents a change from previous practice), and I think all the bilateral relations articles should be brought into line en masse - it's no big deal, and it's only confusing to recommend one thing and do another.--Kotniski (talk) 09:09, 3 August 2011 (UTC)

But this, just a few minutes ago. Tony (talk) 09:16, 3 August 2011 (UTC)

All right, here's a proposal to get this started.--Kotniski (talk) 12:52, 4 August 2011 (UTC)

Chronological items - what should go into MOS and what should go into MOSNUM?

I have reinstated the changes that were reverted by User:Kevin McE. While I agree that that the MOS is not a list of possible difficulties, the list of difficulties serves as an introduction to the relevant section in WP:MOSNUM. This follows the conventions used for the introduction in MOS for the section on Units of Measure.

WP:MOS is becoming unmanageable – the split of WP:MOSNUM and of the section on Geographical terms has helped, but there has always been the danger of conflicting information appears in WP:MOS and WP:MOSNUM which is why I wrote the introduction as a series of difficulties. Martinvl (talk) 09:12, 2 August 2011 (UTC)

Now that this article is unlocked, I made the changes discussed previously - see (Wikipedia talk:Manual of Style/Archive 122#Archiving Policy) and (Wikipedia talk:Manual of Style/Archive 123#Size of WP:MOS). My changes were twice unceremoniously undone without either editor actually being aware of the history behind these changes. Rather than have to go through all the discussions in these two archives, I invite comments from other editors regarding the changes that I made to WP:MOS at about 08:00 on 2 August 2011. Martinvl (talk) 13:17, 2 August 2011 (UTC)

My idea is that each section in WP:MOS should be a summary of the full guideline found in a sub-page, giving the general rules you need to know most often and a link to the “full” guidance for rare exceptions and rules which apply in more specialized situations. This is what I had in mind (though I haven't edited that page in ages, so it may be outdated wrt what WP:MOSNUM says now). ― A. di M.​_plé​^dréachtaí 15:21, 2 August 2011 (UTC)

One of the biggest problems with a summary is "mission creep" - more and more of what is on the subsidiary page get included back into the summary. That is why I prefer the introduction to define the scope of what is being discussed on the subsidiary page rather than to summarise it. This also removes the problem of contradictions being introduced into the text. In the case of chronological items, I reduced the scope to a number of typical issues that the user might have (and issues that are relevant but that he had not though of). Martinvl (talk) 15:28, 2 August 2011 (UTC)

I reverted Martin's edit again. I'm not opposed, but while I see there was some support for this kind of thing a few weeks ago, it's not clear to me that this was the wording we wanted for the summary. Are there any objections to reducing the verbiage by reinstating Martin's edit? — kwami (talk) 18:48, 2 August 2011 (UTC)

A Manual of Style needs to be a manual, something that tells people what to do and how to do it. A list of non-directive principals (sic) is not a manual. It is nonsensical to list what a manual clarifies, while excising those clarifications. So long as the section headings are clear and accurate, so that the specific issue can be navigated to quickly via the ToC, why does the length of a MoS page matter any? Kevin McE (talk) 19:50, 2 August 2011 (UTC)

The Manual of Style comprises lots of pages, this one is only the “front” page. If all the guidance was found unabridged on one page, it'd be 1.4 MB, which would take forever to download with slower connections and would crash the readers' browsers a sizeable fraction of the times. :-) My idea is having the bulk of the guidance in the subpages (each of which would roughly correspond to a chapter of a printed style guide), a front page at WP:MOS giving an introduction with the general principles and then a summary of each subpage, and maybe a “full” version, for “brave” readers with a fast connection and no fear of a ToC filling several screenfuls, which would have the same introduction and then transclude the subpages. ― A. di M.​_plé​^dréachtaí 21:16, 2 August 2011 (UTC)

Agree with A. di M. "So long as the section headings are clear and accurate, so that the specific issue can be navigated to quickly via the ToC" It can't. Most of the subpage information can't be found that way. But in the case of date information, you can still find it through the ToC, after clicking the link. That's better than finding a duplicate section that may or may not match the MOSNUM information, specifically contradictions number 9 and 10, and any other contradictions that may accumulate in the same way. Yes there are typos; we can fix them. And the length of the MoS page matters because it gives people less information to look through, before concluding that what they really want is either missing or on some subpage they can't find. We need a better system for navigating the subpages, but this step is in the right direction. Art LaPella (talk) 23:14, 2 August 2011 (UTC)

I proposed more than a year ago that MOSNUM be reduced to just the stuff one doesn't want here (radiocarbon dating, for example, which is still on this main page, I see); and that MoS central have the basics as now. Tony (talk) 01:45, 3 August 2011 (UTC)

Are people happy for me to reinstate the summary that was reverted yesterday (2 AUgust 2011)?

Are people happy that I should replace the section on numbers with a similar summary (currently both summaries can be seen at User:Martinvl/MOSNUM)?

Martinvl (talk) 06:52, 3 August 2011 (UTC)

I don't think I am. Can you comment on my previous post here? Tony (talk) 06:55, 3 August 2011 (UTC)

Tony, if you scroll down to the section on Chronological items in [the version that was undone], you will see my proposal. It does not explicitly mention radiocarbon dating, but implicitly accomodates it with the phrase "but not limited to". Is this the answer that you were looking for, or should I have been looking at some other comment that you made? Martinvl (talk) 08:18, 3 August 2011 (UTC)

OK, I may have underestimated the scale of things above, and maybe the section should be abbreviated, but other sections are prescriptive, and Martinvl's text is descriptive: that does not serve as a summary of the MoS. Kevin McE (talk) 08:39, 3 August 2011 (UTC)

Do you mean it would contain exceptions but not the rules they are exceptions to? I think that would be confusing. ― A. di M.​_plé​^dréachtaí 15:20, 3 August 2011 (UTC)

??? I've said nothing about exceptions. If a short version of the MoS is required for the Mos page, then let that be a manual, not a description of the issues dealt with by the Manual elsewhere. Kevin McE (talk) 17:43, 3 August 2011 (UTC)

Actually, I was commenting on Tony's suggestion “that MOSNUM be reduced to just the stuff one doesn't want here”. ― A. di M.​_plé​^dréachtaí 21:26, 3 August 2011 (UTC)

In response to User:Kevin McE - What you would have written, taking note that everything in the summary has to be in the subsidiary article and nothing in the summary and subsidary article should contradict each other. For the record, I cannot accept an answer "People must be careful to ensure that this does not happen" - my real life work experience has taught me that such demands are fraught with danger.

Kevin McE has commented that my proposed rewrite, unlike most of the article, is not prescriptive. I know - the fault lies in the rest of the article - this is the third section to be contracted in a non-prescritptive manner, the other two being "geographical items" and "units of measure". The text for the fourth section "Numbers" can be seen User:Martinvl/MOSNUM. Since I am not a linguist, I will encourage other editors to use these four sections a start point for contracting the other sections of WP:MOS. Martinvl (talk) 09:02, 3 August 2011 (UTC)

the fault lies in the rest of the article: is this your opinion, or is there a consensus that mandates this fundamental change? Kevin McE (talk) 10:15, 3 August 2011 (UTC)

This is essentailly my opinion, based, I believe, on sound experience in real life. It was good enough for the user community when I applied it to the section "Units of measure", so it should still be good enough. The user community is generally of the view that something is needed, buit nobody is too sure what is needed (see (Wikipedia talk:Manual of Style/Archive 122#Archiving Policy and (Wikipedia talk:Manual of Style/Archive 123#Size of WP:MOS). As nobody else has done anything positive about this particular problem, I am trying to be WP:BOLD, putting a solution in place and seeing if the user community likes it. Martinvl (talk) 13:45, 3 August 2011 (UTC)

I think that I have answered all the queries concerning this change that I originally made a few days ago and that was subsequently revoked. Both editors who revoked it now understand my rationale and in the absence of any further comments, negative or positive, I am reinstating it. Martinvl (talk) 13:57, 4 August 2011 (UTC)

Punctuation after date

Neither WP:MOS nor WP:MOSNUM appear to define whether a comma is required after the year in cases like the following:

• They agreed a ceasefire on April 1, 1873 in order to celebrate April Fools' Day.

or

• They agreed a ceasefire on April 1, 1873, in order to celebrate April Fools' Day.

Which is correct? All examples given for this date format seem to have something like a full stop or right parenthesis after. --Stfg (talk) 10:04, 17 July 2011 (UTC)

• MLA Handbook for Writers of Research Papers, 7th ed., §3.5.6: "If you begin with the month, be sure to add a comma after the day and also after the year, unless another punctuation mark goes there..."

• Associated Press Stylebook (2007), p. 159 (months entry): "Feb. 14, 1987, was the target date." Jc3s5h (talk) 11:22, 17 July 2011 (UTC)

• The comma after date is placed when the date if part of an introductory clause, i.e. "In July 2011, the weather was nice." or "As on July 2011, I will stop drinking." However the first clause in the example you give is not an introductory clause, it is just a regular clause that happens to have a date in it. You could replace it with, say, "They agreed a ceasefire yesterday in order to celebrate April Fools' Day." In this case "in order to celebrate April Fools' Day" is not an independent clause, so a comma is not needed. —  HELLKNOWZ  ▎TALK 11:56, 17 July 2011 (UTC)

• Of course, H3llknOwz's opinion conflicts with the two cited style manuals. Jc3s5h (talk) 12:18, 17 July 2011 (UTC)

• No, it doesn't. It is not opinion, I am referring to English language punctuation of introductory phrases. Also, I didn't say you cannot use a comma, I said it is not needed. Both your examples refer to introductory phrases – "If you begin with the month" and "Feb. 14, 1987, was the target date." are both used at the start of sentence. What you are possibly meaning to quote is MLA §3.2.2 "If such a date [month, day, year] comes in the middle of a sentence, include a comma after the year." This is a stylebook suggestion, and whether it is used or not on Wikipedia is at editor discretion/consensus. I did not say you cannot use it and I believe most editors do. —  HELLKNOWZ  ▎TALK 13:12, 17 July 2011 (UTC)

• The statement from the MLA handbook was an instruction, not an example. Failure to put a comma or other punctuation after the year is an error in applying MLA style. Of course, as you say, any style book only applies to those who choose to follow it. Jc3s5h (talk) 13:46, 17 July 2011 (UTC)

Thank you both for helping with this. I've looked elsewhere on the net and it does appear more common to use the comma. But it is a style question, so the real issue is whether this should be left to individual editors or whether MOS should have a position on it. --Stfg (talk) 13:21, 17 July 2011 (UTC)

Chicago (14th ed.) says "In the alternative style [i.e. when you begin with the month], however, commas must be used before and after the year."--Boson (talk) 13:41, 17 July 2011 (UTC)

I have had discussion on this topic with other editors. I am not saying that these discussions are definitive, as I am sure there are plenty of style books, but feel free to read what was said. User:Chaosdruid/usefullinks/GOCEconv#Year and comma Chaosdruid (talk) 14:48, 17 July 2011 (UTC)

Hellknows' post is correct, if a little unclear. If there is a comma before the year, there must be a punctuation mark after it (usually a comma, or the period after the sentence); the year is parenthetical. If there is no comma before, there may or may not be a following comma, depending on the rest of the sentence. In July 11, 2011, there must be a comma before; in July 2011 or 11 July 2011 there isn't. Septentrionalis PMAnderson 19:24, 17 July 2011 (UTC)

The rule isn't in the Manual of Style, but it is at WP:COPYEDIT, and I cite it so often it needs a shortcut: "... the month day, year, style of writing dates punctuates after the year ..." Art LaPella (talk) 20:21, 17 July 2011 (UTC)

The discussion on Chaosdruid's talk relates to sentences beginning In 2011, the.... That is a short parenthetical phrase, and it is a question of taste whether it needs to be set off, by a single comma, at all. But when you use one comma in July 11, 2001, it is mandatory to use both. Septentrionalis PMAnderson 20:47, 18 July 2011 (UTC)

As Hellknowz and Anderson said. Think of it as a parenthetical: They agreed a ceasefire on April 1 (1873) in order to celebrate April Fools' Day. You couldn't leave out the closing parenthesis. That's the rational for requiring two commas, and two commas is standard punctuation. That said, it is extremely common to leave out the second comma, because the main reason they are needed is to help separate the numerals in the day from those in the year, and the clause is perfectly legible with a single comma between the day and year. I would therefore say that in a colloquial context a single comma is perfectly acceptable. However, in a formal context, such as an encyclopedia, we should stick to formal punctuation. — kwami (talk) 03:59, 19 July 2011 (UTC)

I don't think this additional comma should be the subject of MoS rules, as long as there's article consistency on the matter. Aside from this, both alternatives at the top of this thread are unsatisfactory. "They agreed to a ceasefire on April 1, 1873 in order to celebrate April Fools' Day." Tony (talk) 04:32, 19 July 2011 (UTC)

"They agreed a ceasefire" is grammatical, but apparently only in UK and Irish English; see wikt:agree#Verb. I didn't know that, but I am British and this would be the natural, though not the only, way for me to say/write it. The "in order" part of "in order to" is certainly redundant, but that doesn't make it actually wrong – I'm not a minimalist :) --Stfg (talk) 10:08, 23 July 2011 (UTC)

Indeed. ― A. di M.​_plé​^dréachtaí 14:55, 24 July 2011 (UTC)

(Chuckle) A classic worthy of The Daily Show or The Colbert Report. --Jackftwist (talk) 21:23, 29 July 2011 (UTC)

Re "in order to": Yeah, it is indeed redundant (or should that be superfluous?), at least in a huge majority of cases. But it also sometimes makes the sentence read more smoothly. I do omit it wherever possible. But I couldn't tell you how many of those times I've later reread the sentence, especially after more than a few minutes elapsed, and found myself stumbling over an awkward lack of transition between the infinitive ("to ...") and what preceded it. It causes me to pause for a few moments, thinking a word or more has been left out. And that's in something I wrote (or at least edited) myself, so I presumably know what I meant to say! If the absence of "in order" causes me to stumble in those cases, might not someone else, reading the sentence for the first time, also stumble? And that's an absolute no-no in effective writing. As Michael Oreskes, senior managing editor of the Associated Press, said, "The faster the eye flits across the words, the more vital it is that language be immediately and abundantly clear." (See

http://www.washingtonpost.com/lifestyle/style/david-minthorn-is-the-grammar-expert-for-the-associated-press/2011/07/25/gIQAGBLwfI_story.html

p. 2, 4th paragraph from the end.) So I usually do omit "in order," but I try to be sure what's left is "immediately and abundantly clear," with no stumbles. --Jackftwist (talk) 21:23, 29 July 2011 (UTC)

That comma rule's history at WT:COPYEDIT has ranged from wanting to delete the rule because it's wrong to wanting to delete it because the rule is so well-known we don't need it. As I stated there, if we didn't have the rule written down somewhere, then enforcing it would get me into a crossfire between those who consider the rule to be obvious vs. ridiculous, and I would probably decline to enforce it. WP:COPYEDIT is no longer considered part of the Manual of Style, but so far that hasn't mattered when discussing a particular comma. Art LaPella (talk) 04:59, 19 July 2011 (UTC)

• Without suggesting a view as to whether or not this should be part of MOS rules, I will say that personally -- based on whatever guidebooks I've read over the decades -- I do insert a comma after the year, and do view that as appropriate.--Epeefleche (talk) 20:25, 19 July 2011 (UTC)

As regarding the parentheses comment, the correct application could also, and possibly more correctly, be seen as:

They agreed to a ceasefire (on April 1, 1873) in order to celebrate April Fools' Day. or

They agreed to a ceasefire on April 1 (1873), in order to celebrate April Fools' Day.

The other possibility:

On April 1, 1873, they agreed to a ceasefire, in order to celebrate April Fools' Day. - I do not think this is right though, as I firmly believe that should be written as:

On April 1, 1873 they agreed to a ceasefire, in order to celebrate April Fools' Day. There is an obvious ambiguity as to whether the ceasefire, or the agreement, was the 1873 date. In order to make it more plain that the agreement and the date of ceasefire were the same, I think the date would have to come first; if we use the first set of options the date could be either.

I realise this is only an example, but it is an somewhat minor(?) important point. Chaosdruid (talk) 11:55, 5 August 2011 (UTC)

Extraneous clutter at the top

I haven't read the opening paragraphs of the MoS for some time, and was perplexed to see this:

See also: Manual of Style:Trademarks and Wikipedia's policy on no original research

Many points of usage, such as the treatment of proper names, can be decided by observing the style adopted by high-quality sources when considering a stylistic question. Unless there is a clear reason to do otherwise, follow the usage of reliable English-language secondary sources. If the sources can be shown to be unrepresentative of current English usage, follow current English usage instead—and consult more sources.

This is counter to the opening statement "If the Manual of Style does not specify a preferred usage, please discuss the issue on the talk page." It seems to undermine the whole function of a MoS. There's an editorial note within the paragraph saying that there's no consensus for the opening statement. The links below the title are weird. I cannot find a record of consensus for adding this stuff.

Unless someone can provide a good reason for retaining it, I intend to remove this subsection in the interests of reducing clutter and redundancy on the sprawling MoS page (especially since it's confusing). Tony (talk) 04:18, 3 August 2011 (UTC)

Here's the edit. It's two years old. Its author, who often uses that edit's same reasoning to this day, has been banned for a week. We may thus presume a difference of opinion will occur next week. Art LaPella (talk) 04:50, 3 August 2011 (UTC)

We don't need to wait for PMA's opinion, as we have it on record here: Wikipedia_talk:Manual_of_Style/Archive_123#Alternative_proposal:_just_remove_it; he would support removing it under the condition that MOS is no longer a guideline; that is, in his mini manifesto, he objects to the removal and objects to MOS in general, and objects to anyone who supports MOS as "regulars" and "sponsors" who "need bots" (I'm still puzzling over that last bit). Dicklyon (talk) 07:32, 3 August 2011 (UTC)

Ah, fine, but this transcends a one-week ban. Tony (talk) 05:05, 3 August 2011 (UTC)

Support. I agree with revisiting this. This is the WP MOS and for stylistic issues we should offer the first port-of-call for our editors. If there's something we don't cover, then it can be discussed on the relevant talk page. The default should not be to go for whatever external manual an editor believes appropriate—that way leads to edit wars. GFHandel ♬  05:18, 3 August 2011 (UTC)

Support deletion of "follow the sources". Yes, it's a bad section; we had already been discussing deleting it, when the MOS got locked for other disturbances. See Wikipedia_talk:Manual_of_Style/Archive_123#Alternative_proposal:_just_remove_it, where there was considerable support for removing it. History wise, in the original discussion about PMA's "follow the sources" section that met some initial rejection, several expressed concern that it would conflict with our style guidelines: http://en.wikipedia.org/wiki/Wikipedia_talk:Manual_of_Style/Archive_109#Follow_the_sources PMA "persuaded" them that it would be OK, and then proceeded to use it exactly as had been feared. This will continue (next week) if we don't get rid of it. Dicklyon (talk) 05:39, 3 August 2011 (UTC)

Support. This was a politically motivated and unilateral insertion from the start. Its aim was to relegate WP:MOS, under the banner of Motherhood, to a minor role among the Project's guidelines and policies. The addition was immediately reverted (see edits following the one Art LaPella has linked, above). It attracted no thorough analysis in discussion (only two or three editors were involved), and never had anything remotely resembling consensus. The section's inclusion has been the cause of months of protection at this page, earlier this year. Time for it to go. No one denies that reliable sources have their place; but let no one deny the role of the Manual of Style. A manual of style is not an inventory of aphorisms; nor is it a vehicle for apologising for its own existence, at every available opportunity. Noetica^Tea? 05:44, 3 August 2011 (UTC)

Support. Oh, yes please, with a cherry on top! It's useless waffle that doesn't add anything, and creates ambiguity for the average user when it's certainty they seek when coming here. The MOS needs to lose some calories, and this will burn up a few. --Ohconfucius ^¡digame! 06:31, 3 August 2011 (UTC)

Comment I like the idea of a section reminding editors to consult sources because the MoS will not cover all instances or issues. However, it could be better phrased and I would not object to removing this section altogether for now. We can always put it back if we need it. Also: 1. I don't think that's quite what PMA meant, but it isn't as if we need a quorum or any specific editor's permission to change any part of the MoS. 2. The MoS register was a nice experiment, but it hasn't gone anywhere. Zero objection to giving it an honorable burial. Darkfrog24 (talk) 12:28, 3 August 2011 (UTC)

If (and that's a big if) there is instruction to consult outside sources, it should be clear that should only happen if the WP MOS does not give instruction and nothing has been forthcoming after discussion on the relevant MOS talk page. There should also be guidance to introduce discussion about what was discovered on the outside sources. I thought about why the MOS is important to WP and came up with the following (and I'm sure the list isn't exhaustive):

• it addresses the peculiarities of having a vast number of different viewing devices (different sized screens, tablets, PDAs, etc). This is something that print-based MOSs don't have to consider.
• it considers features specific to WP (e.g. info-boxes, tables, links, media, dynamic formatting, etc).
• it provides consistency for the reader across millions of articles (e.g. to avoid encountering different abbreviations, or units of measure formatted differently in different articles).
• it indicates how information can be expressed succinctly—allowing the reader to focus on detail that deepens the understand of the article.
• it gives guidance to local editors (to avoid or settle style-based disputes—encouraging to get on with the business of adding content).
• it gives guidance to editors who run tidy-up scripts (across multiple articles). Consistency also means that scripts have a better chance of working reliably (e.g. a smaller set of date formats exist to recognise and parse).

What's amazing about WP's MOS is that it is one of the greatest exercise in consensus imaginable: with (just the main MOS page having) over 7,500 edits in ten years by more than 2,100 distinct editors (and five times that many edits on the talk page) [6][7]. No doubt we are still on a (hopefully asymptotic) journey, but congratulations to all the editors here who have worked hard to achieve such noble goals.

GFHandel ♬  00:14, 4 August 2011 (UTC)

• Support. Seems reasonable, and from the above has consensus support.--Epeefleche (talk) 01:13, 4 August 2011 (UTC)

• Agree Either tone it down, saying "when the MOS doesn't offer guidance, it might be a good idea to do what others do", or get rid of it completely. See Dicklyon's comments for why. Headbomb {talk / contribs / physics / books} 15:57, 4 August 2011 (UTC)

The global view section

So what the hell is this doing there? I don't mind "principles", but some of them seem irrelevant and/r redundant. No wonder editors complain that the MoS is too large:

"Except in content with a local focus or where specific localized grammar or spelling is appropriate, or when a precedent has been established and no clear reason has been accepted by a consensus to overturn it, content should be presented from a global view without bias towards any particular culture or group." Tony (talk) 05:10, 3 August 2011 (UTC)

Tony, that section came in a year ago, in this diff by Doc Quintana (talk · contribs · deleted contribs · logs · edit filter log · block user · block log), who was subsequently permanently blocked as an abusive sock puppeter, with summary "I thought I saw this somewhere", and little notice. Given the fact that nobody has discussed it, and it seems extraordinarily odd, and has no better justification than that he though he had seen it some place, I suggest we just take it out, and see if anyone cares. I'll do that. Dicklyon (talk) 06:00, 3 August 2011 (UTC)

And the "general principles" section that these were stuck into had its start in this diff by SlimVirgin. Probably, in retrospect, saying "General Principles" made it a magnet for people with an agenda to sneak things into. Dicklyon (talk) 06:13, 3 August 2011 (UTC)

Well, Slim's good with this stuff, but probably didn't envisage that it would become the unruly forest we now see. Tony (talk) 06:40, 3 August 2011 (UTC)

The analogy applies inasmuch as the need to prune from time to time. --Ohconfucius ^¡digame! 07:09, 3 August 2011 (UTC)

It's like "list" sections that become magnets for trivia, against the best intentions of their creators. So yes, let's reorganize it in a way that won't attract more principles. Or go with your forest pruning if you don't like my magnetic metaphor. Dicklyon (talk) 07:21, 3 August 2011 (UTC)

Two more things: I can't imagine who might want to use any of those shortcuts to the "principles", or where. Isn't it easier and simpler to conflate all into a readable whole and refer people to the lead of the MoS? And could I ask whether Wikipedia:Manual_of_Style/Register is worth pursuing; it looks like a ghost town, yet is responsible for a large clutter-box at the top of the MoS. Tony (talk) 07:59, 3 August 2011 (UTC)

Your imagination is lacking. PMA used to cite MOS:FOLLOW some times when he disagreed with the application of the MOS. Dicklyon (talk) 01:32, 4 August 2011 (UTC)

Essentially, that is WP:COMMONALITY and WP:WORLDVIEW. The former is already in the MoS (though buried somewhere most readers won't notice) and the latter is more about content than style per se. ― A. di M.​_plé​^dréachtaí 13:02, 3 August 2011 (UTC)

Good job, Tony; the new lead without the "general principals" collection is much better. Dicklyon (talk) 01:30, 4 August 2011 (UTC)

• Yes, it works for me. Greg L (talk) 01:48, 4 August 2011 (UTC)

With Noetica's copy-edits for conciseness and such, we've dropped the size by about 4500 bytes (about 3%); not huge, but a good counter to the recent creep. What else can be cleaned up? Dicklyon (talk) 05:45, 4 August 2011 (UTC)

Yes, it appears that User:Noetica is going through the MoS (in reverse order?) performing much-needed maintenance. I expect he'll alert the talk page if there's anything substantive to change. It looks like it's approaching a 10% reduction, but I'd be happy with a more minimalist approach (although my own effort a year ago showed that it's possible to make reductions in the order of 60%, I think people wouldn't be happy with the small number of examples that entails). Tony (talk) 06:15, 4 August 2011 (UTC)

Reducing this page to principles only

I haven't been following MoS discussions recently, so this sort of thing has probably been proposed before, but how about making this main MoS page contain only general principles and perhaps a few of the most frequently-arising style issues, and moving all of the detailed guidance onto subpages (as a lot of it is already). For example, we could move all the complex detail about dashes and other punctuation marks to WP:Manual of Style (punctuation). Thoughts?--Kotniski (talk) 14:39, 4 August 2011 (UTC)

If we execute the previously approved process of moving all the subsidiary pages into organized subpages (like WP:Manual of Style/punctuation or such), then we could put here a good "table of contents"-like set of mini sections with main links. So more than principles, but a good portal. It will be a lot of work, but it seems like we're working well together now, so I don't see why we can't find a good path and get it done. But I don't think "principles only" is exactly the right model, given the trouble we've seen with the "general principles" section accumulating odd things. Dicklyon (talk) 14:52, 4 August 2011 (UTC)

There is a general consensus that this needs to be done, but there has been debate as to how it should be done. I have already done the sections on units of measure (a few months ago) and on chronological items (today). I have the text ready for the section on numbers which I will insert as soon as the section on chronological items has been accepted. Somebody else did the section on geographical items (the difference in style should be obvious).

Yes, I would gladly see the section on punctuation being moved to a separate section with an introduction in the main article. The main debate however is what the introduction should consist of - my view is that it should define the scope of the main section (such as the Chronological items section and the Units of measure section, others suggest that it should be a summary in itself (such as the Geographical items section). Martinvl (talk) 14:56, 4 August 2011 (UTC)

Bad idea. That would force editors to go to six or seven different pages to find the information they need. If it's all on one big page, all they have to do is hit CTRL-F. Darkfrog24 (talk) 15:03, 4 August 2011 (UTC)

CTRL-F is fine if you know what you are looking for. If however you are browsing to get a feel as to what is there, a huge page is IMO unworkable - I prefer to go to subsidiary pages. Martinvl (talk) 15:10, 4 August 2011 (UTC)

I mean, it already decidedly isn't all on one page - such a page would be too big to load. It doesn't seem to make much sense to move most of the detail off to subpages, while leaving a few parts of the detail on the main page. If we just left the really important things here, people could more easily find those things, and even have their attention drawn to them despite their not specifically looking for them.--Kotniski (talk) 15:14, 4 August 2011 (UTC)

I agree. While this view may appear to make sense from a philosophical perspective, we shouldn't make editors work harder than they need to in order to adopt these guidelines. There's already an ever increasing learning curve for new editors, and adding layers of links for lookups only makes it more difficult. Likewise, I shouldn't have to waste extra time drilling down to find the guidelines I need. Thanks. Regards, RJH (talk) 15:51, 4 August 2011 (UTC)

Bad idea. The current summary style MOS mainpage is perfectly adequate in scope (well baring the occasional prunings). Headbomb {talk / contribs / physics / books} 16:00, 4 August 2011 (UTC)

"...baring the occasional prunings..." sounds like the fifteen-year-old who was "a little bit pregnant". Martinvl (talk) 16:43, 4 August 2011 (UTC)

These comments seem to be based on the assumption that the entire MoS is currently on the main page, and that we're proposing breaking it up in order to make things harder to find. But that isn't the case at all - the vast majority of it is on separate pages already, I just don't understand why we have to have a few things (like punctuation) left on the main page in great detail, because no-one's yet got round to creating a subpage for them. No doubt you'll say that it's because punctuation is something that applies to all articles whatever the topic area – but the same is true of dates, and that's all been farmed off to a subpage. We should accept that it's quite unrealistic to put our entire style manual on one page; consequently, we should use this central page as a good introduction and navigational hub for the manual, not cluttering it up with a fairly random selection of the detailed rules.--Kotniski (talk) 17:01, 4 August 2011 (UTC)

Actually, what I wouldn't mind seeing is a complete MoS in a PDF file so I can perform rapid searches in a separate window. (Or a dynamic, searchable web page.) Regards, RJH (talk) 17:45, 4 August 2011 (UTC)

The MoS should not be made more decentralized than it already is (and preferably should be made less so where appropriate). The main MoS should contain all the things that an editor might expect to find in any other style guide: things like punctuation would fall into that category. Only specialized content, things that would be relevant only to a small number of articles, should be considered for separate pages. Things like heraldry and chemical formulae might fall into that category.

In addition to searchability, there is another advantage to having all or most of the MoS on one page: It allows the reader to tell about how much content there is. When one has to click from one page to the next to the next to the next there's never any end in sight. It's depressing. Darkfrog24 (talk) 17:59, 4 August 2011 (UTC)

Many of the so-called summaries that we have in WP:MOS are 50% of the length of the corresponding subsidiary article. The normal advice is that an introduction should be no more than 10% of the entire article. This suggests to me that the so-called summary is far too long. [The section on punctuation is the longest section that has no subsidiary article.[User:Kotniski|Kotniski's]] suggestion that there be a new article dealing with punctuation merely proposing that we continue with the process that is in place.

Recently the whole of the MoS was locked for several weeks because of an edit-war concerning dashes. If we continue to break WP:MOS up, the impact of locking the MoS will affect far fewer users.

RJH suggested that the MoS be availalbe as a PDF file. This has already been proposed. The mechanism will be to create a Wikibook which is a concatenation of all the articles in the MoS. Of course somebody will have to sit down and do some donkey work.Martinvl (talk) 19:05, 4 August 2011 (UTC)

The edit lock affected us, not the editors consulting the MoS for use. It is right and proper to inconvenience the talk page regulars if it means that the users get a better experience. The matter could also be solved by finding a way to edit lock individual sections of the MoS while leaving other accessible.

If the summary is 50% as long as the subpage, then we might as well ask why have the subpage. Moving material onto the main page would eliminate redundancy just as well. Darkfrog24 (talk) 20:00, 4 August 2011 (UTC)

To make Darkfrog's idea work, we would need a way for readers to decide if their topic is more like punctuation or more like heraldry. Readers would need to learn two separate searching strategies, and choose the strategy that fits the answer to that question. I'm pretty sure there is no way to expect readers to guess how we have divided topics into punctuation-like topics and heraldry-like topics. Art LaPella (talk) 22:01, 4 August 2011 (UTC)

But more important than finding the best reorganization plan, is to actually do something this time. Even regulars are unaware of about half the rules, in part because they are unnecessarily hard to find. And yet we debate things like dash spacing as if changing our contradictory rules is the same as changing Wikipedia. Art LaPella (talk) 22:10, 4 August 2011 (UTC)

(edit conflict) @Headbomb: I agree (at least with the principle, though not necessarily with every detail of the current implementation), but there are a few sections (namely “Quotations”, “Punctuation” and “Grammar”) which don't use summary style, but give the unabridged ‘full-blown’ guidance in the MOS front page with no summary or overview of them anywhere. I think it wouldn't be a bad idea to split them to subpages of their own and retaining a summary/overview on the MOS main page. In fact, I think that even if in a first moment after the split they were transcluded on WP:MOS until summaries are completed, that would have a positive effect as discussion specifically about one of them would be on their own talk pages rather than here. Since this page is more than half a friggin' megabyte right now even with threads being automatically archived after one week (which causes my browser to fail loading it all about 50% of the times), I would gladly welcome such a change. ― A. di M.​_plé​^dréachtaí 22:16, 4 August 2011 (UTC)

Current state of affairs

• In some cases there's a one-to-one correspondence between a WP:MOS section and a subpage:
  • Wikipedia:Manual of Style (capital letters) (22,004 bytes); WP:MOS#Capital letters (12,203 B, 55%);
  • Wikipedia:Manual of Style (abbreviations) (18,810 B); WP:MOS#Abbreviations (6036 B, 32%);
  • Wikipedia:Manual of Style (linking) (39,728 B); WP:MOS#Links (2136 B, 5%).
• In some cases a WP:MOS section corresponds to part of a subpage:
  • WP:MOSTEXT#Italic face (10,224 B); WP:MOS#Italics (4705 B, 46%)
  • WP:MOSNUM#Chronological items (23,852 B); WP:MOS#Chronological items (1153 B, 5%)
  • WP:MOSNUM#Numbers (18,514 B); WP:MOS#Numbers (5722 B, 31%)
  • WP:MOSNUM#Currencies (4049 B); WP:MOS#Currencies (1319 B, 33%)
  • WP:MOSNUM#Units of measurement (27,825 B); WP:MOS#Units of measurement (1897 B, 7%)
• There are weird things:
  • WP:MOSNUM#Common mathematical symbols (2770 B) and WP:MOS#Common mathematical symbols (1973 B) partly overlap but neither is a subset of the other, and both mention WP:MOSMATH which however (surprisingly) doesn't cover all of that stuff.
• A few WP:MOS sections have no corresponding subpage, and a lot subpages have no corresponding WP:MOS section.

Note that the percentages don't correlate perfectly with informativeness: WP:MOS#Units of measurement gives no actual concrete recommendation whatsoever, but WP:MOS#Links, even though its size is a smaller fraction of that of the full guideline, does. ― A. di M.​_plé​^dréachtaí 23:42, 4 August 2011 (UTC)

FWIW, my idea would be that WP:MOS, except for the lead and possibly a section "General principles" at the top and/or a section "Miscellaneous" at the bottom, entirely consisted of summaries/overviews of subpages, with each H2 section corresponding to one subpage; that the threshold of inclusion for each particular point in WP:MOS would be “Is this relevant to most articles?” (which would mean that some WP:MOS sections such as Grammar or Punctuation would be 40% of the subpage, other sections less than that, and some subpages such as WP:LDSMOS would have no corresponding section at WP:MOS at all, though there could be a list of such subpages in WP:MOS#Miscellaneous); and that there might be a WP:Manual of Style/Full page transcluding all the subpages, so that one could search with Ctrl-F. ― A. di M.​_plé​^dréachtaí 23:57, 4 August 2011 (UTC)

Yeah. Something like that. Is it true that renaming the subpages makes the Table of Contents list everything? That would be an additional way to find stuff. Art LaPella (talk) 00:21, 5 August 2011 (UTC)

I'm struggling with "one to one" versus "part". Tony (talk) 02:08, 5 August 2011 (UTC)

Not per se, but headings in transcluded pages as in this hypothetical WP:Manual of Style/Full do show up in the ToC, regardless of their title. ― A. di M.​_plé​^dréachtaí 04:18, 5 August 2011 (UTC)

I agree with the principle that the sections of the main page should be overviews of the subpages (that was probably what I had in mind with my proposal above), but with the focus on the points that people are most likely to be wondering about. And since the philosophy behind the MoS is mostly about consistency (where there's no good reason to be inconsistent), we should set a good example by having some consistency on our own pages. --Kotniski (talk) 19:24, 5 August 2011 (UTC)