Wikipedia talk:Manual of Style/dash drafting/discussion

This page is for discussion related to the poll taking place at WT:Manual of Style/dash drafting.

I am setting this up to remove discussion material and declutter the parent page. We really need to keep the parent page navigable and on topic. I will be moving material here soon. Casliber (talk · contribs) 21:15, 2 June 2011 (UTC)

---

Abbreviations used on this page

---

What do we agree upon?

I have included my own opinions so we can see where we disagree. If we all disagree with the same point in the same direction, we may be able to find a new point where we agree; otherwise our starting point should be what hardly anybody disagrees with.

I acknowledge those who want only hyphens and those who want exactly this text will agree/disagree with everything; if they comment, therefore, the result will be to strike everything. Septentrionalis PMAnderson 03:34, 19 May 2011 (UTC)

(It is a pity this page doesn't have a talk talk page really) PMAnderson do the following items cover all areas of the dispute? You've done the right thing in succinctly itemizing each use. We need to ensure the page remains easy to navigate and gage consensus. Casliber (talk · contribs) 04:21, 19 May 2011 (UTC)

There are some related issues with WP:HYPHEN, but they may fall out if the dashes are settled.

Unfortunately, I'm going to have to change my mind about this; one section called for. Septentrionalis PMAnderson 22:26, 19 May 2011 (UTC)

There are also more general issues on what MOS is, most of them mentioned in the RfAr; but they aren't really covered in the MOS text, so this approach won't work. Septentrionalis PMAnderson 04:28, 19 May 2011 (UTC)

I think it would be a great opportunity while this is the focus of the wider community to examine usage of a related area such as WP:HYPHEN but am not familiar with them so if you reckon they will settle spontaneously that sounds prudent. Agree the broader issue of the strength/weight MOS actually has is a bigger issue and arguably beyond the scope of this. Casliber (talk · contribs) 04:33, 19 May 2011 (UTC)

I don't think we can really address what led to the arbcom thing and all the fuss about this unless we do address the strength/weight of the MOS. IOW, assuming the style guide is not amended to prefer hyphens in this case (hypothetical) will I need to open a move request to move from Carbon-carbon bond to Carbon–carbon bond where we can repeat this debate and see who happens to show up? It would be nice to not have to have this same discussion over and over, with different outcomes depending on what is on whose watchlist. ErikHaugen (talk | contribs) 17:14, 19 May 2011 (UTC)

I could not agree more. To amplify, absent consensus for changing the MOS, the page moves to dashes would seem in compliance. Depending, of course, on what the MOS means. JeffConrad (talk) 20:17, 19 May 2011 (UTC)

The present text of MOS does not have consensus now. This position is absolutely unacceptable; to insist that a text on which we are divided authorizes those who wrote it to do whatever they want until they admit there is consensus against them is exactly what has been wrong with MOS for five years. Septentrionalis PMAnderson 20:43, 19 May 2011 (UTC)

Also this page is intended to judge unanimity, not consensus; we already have an RFC which shows that this text isn't consensus. On the other hand, a next limited to what everybody agrees on will still draw objections from those who wanted to include something else. Septentrionalis PMAnderson 04:40, 19 May 2011 (UTC)

This well may be your intent, but that′s not how I read the ArbCom motion

Interested parties are instructed to spend the next 14 days from the passing of this motion . . .

which directs us to

1. determining the structure of a discussion on En dashes in article titles to obtain consensus.
2. gathering of consensus on the issue. The discussion should be of sufficient structure to allow easy quantification of consensus rather than a large amount of poorly-framed debate.

It seems to me that we’ve largely skipped 1, and may be taking a different approach to 2 than was given. To be honest, I have no idea what was meant by the the final sentence:

If after two months, a determination isn't realised, a case will be opened and conduct violations will be dealt with severely.

JeffConrad (talk) 20:27, 19 May 2011 (UTC)

I created this page, in response to a suggestion that we determine unanimity. It is not intended to produce a consensus text; it is intended to inform us what would be a minimal text, containing only points to which almost everybody agrees. That text may or may not be consensus; in either case, this is only a first step. Septentrionalis PMAnderson 20:43, 19 May 2011 (UTC)

Concede that this is perhaps an initial step, assuming we agree on the structure of the discussion. JeffConrad (talk) 20:36, 19 May 2011 (UTC)

I doubt ArbCom will think well of anybody who argues about the shape of the table. Please don't. Septentrionalis PMAnderson 20:43, 19 May 2011 (UTC)

• Comment: "if they comment ... the result will be to strike everything". If I may translate into Kwamese (since I'm well known to not speak "English"), it sounds like this means that Anderson thinks he owns this page, and will strike out opinions that do not approximate his own. Do I understand that right? We will then arrive at "unanimity" by disregarding those opinions which diverge from it, a rather Orwellian definition of the term. — kwami (talk) 10:18, 19 May 2011 (UTC)

No no, this page is being observed and we'll ensure we get an idea of consensus. Some uses may be similar below but keeping each item as segmented as possible will avoid "yes/no" or otherwise conditional/partial answers to any bits. Casliber (talk · contribs) 10:57, 19 May 2011 (UTC)

• You guys need to make this page easier to find; I barely managed to do so. The 1st RfC has some 34 !votes about equally split for/against. The 2nd one only 14 with about the same ratio split. I don't want to make any predictions here, but people are getting tired... Tijfo098 (talk) 11:01, 19 May 2011 (UTC)

Agree - I'd think centralised discussion, and notifying all voting editors in previous to reiterate their votes here would be good. I just wanted to ensure that all the questions that needed asking have been listed below. Casliber (talk · contribs) 11:32, 19 May 2011 (UTC)

Oh, I thought below was just to figure out which of the numbered items in ENDASH needed discussion. ErikHaugen (talk | contribs) 18:45, 20 May 2011 (UTC)

to/vs.

• The question of whether MOS permits or requires things isn't up for discussion here, but since the question has been raised, MOS is required at FAC, and WP:DASH is being treated as if it's required by some; thus the Arbcom case. The 4–3 win example belongs in the previous subsection. I won't speak to other varieties of English here, but "Lincoln-Douglas debate" is usually written with a hyphen in AmEng. WP:CREDO gives 9 hits, all with hyphens, and see this WT:MOS discussion for what the American style guides say. (I have recommended a dash here previously because I'm aware that other varieties of English are more pro-dash. I didn't want to come across as doggedly pro-AmEng, and as long as it wasn't causing problems, there was value in supporting one easy-to-learn rule. But this page isn't about getting along, it's about reporting accurately.) - Dank (push to talk) 14:22, 19 May 2011 (UTC)

Dank, 4–3 win isn't a range, so it wouldn't fall under 1. 1 never says anything about being restricted to numbers, or that numbers aren't used for other things. — kwami (talk) 02:42, 20 May 2011 (UTC)

Quite right. - Dank (push to talk) 02:46, 20 May 2011 (UTC)

• Generally, I would like the MOS to "require" things where "require" means we can just move pages/edit text without having this same argument thousands of times. Dank - hopefully this is up for discussion? I'm not sure what you mean. I don't think we can realistically ignore this issue, can we? It's been a pretty big part of the recent WT:MOS discussions and the debates leading to them and to the arbcom thing, I think. Maybe we mean different things by "The question of whether MOS permits or requires things"? ErikHaugen (talk | contribs) 16:53, 19 May 2011 (UTC)

  Generally, I agree with Fut. Perf.'s edits in the original discussion here (and in the preceeding section); I very strongly disagree with any effort to make a minority usage compulsory. If we need uniformity on this, we should choose the far more common alternative; but I don't think we do.

  I do not agree with anyone - and I am naming no-one at this point - who holds that "whatever is not compulsory is forbidden". Septentrionalis PMAnderson 20:24, 19 May 2011 (UTC)

  Discussions of how prescriptive we want to get about dashes are fine. I'm just saying that, since several members of Arbcom promised trouts or topic bans for all unless we get the current mess settled in the allotted time frame, I'd rather not throw big questions into the pot that have evaded consensus for years. - Dank (push to talk) 17:22, 19 May 2011 (UTC)

  What then is up for discussion? JeffConrad (talk) 19:33, 19 May 2011 (UTC)

  I don't think anyone is threatening "topic bans for all". My point is just that how prescriptive the MOS is about dashes is a significant part of the current mess. ErikHaugen (talk | contribs) 17:25, 19 May 2011 (UTC)

  That's exactly what Arbcom promised, and the fact that everyone is saying "Well surely they can't mean me, I haven't done anything wrong, they mean the other guys" is exactly why it turned into an Arbcom case. Having said that ... they can't mean me, I haven't done anything wrong :) But I don't want to see anyone else caught by surprise, so please, let's not get too fancy, let's just answer the questions and get to the end of the Arbcom case. When there's no active threat, we'll have time to discuss all of this in depth. - Dank (push to talk) 17:41, 19 May 2011 (UTC)

  No, the motion said: "a case will be opened and conduct violations will be dealt with severely". ErikHaugen (talk | contribs) 17:53, 19 May 2011 (UTC)

Sorry, I meant what some at Arbcom said. My biggest worry is that people are forgetting that this is an Arbcom case, with potential nasty surprises at the end, and certainly for some who won't be expecting it. Here's what Arbcom predicted in the "Arbitrator's opinion" section of that case. I think there's enough reason here to just aim for getting to the end of the case without digressions and without anyone losing editing privileges:

• Iridescent: "such a case is likely to take a huge amount of time, and be very unsatisfactory for anyone involved; WP:RFAR/DDL took five months to come to a decision and resulted in a huge stack of bans, blocks and restrictions for all those involved, including those who initially brought the case. Are you sure this is a firecracker you really want to light?"
• Casliber: "I echo Iridescent in that a case will not be pleasant, and sanctions for aggressive proponents on both sides possible."
• Roger Davies: "In practical terms, we can take a case, which will likely be long and nasty; issue liberal quantities of bans, topic bans and restrictions; and ask the community to forge a solution via an RFC."
• Sir Fozzie: "Does anyone see any resolution of an ArbCom case that wouldn't end up in "Trouts for all" or "Topic Bans for all?"."
• NewYorkBrad: "Much of the user conduct surrounding this issue has been wretched. All editors who have used excessively strident rhetoric in discussing this issue, or have otherwise acted obnoxiously, should stop doing that."
• JClemens: "but I consider the impact of tolerance (or lack thereof) for individual article differences to be a legitimate conduct matter inasmuch it provides a barrier to editing by those unfamiliar with its nuances." ["Sanctionable conduct matter" seems clear to me in context.]
• - Dank (push to talk) 18:49, 19 May 2011 (UTC)

( facilitator note: Current policy for MOS seems to lie about 2/3rds along the way from "Optional" to "Required". Each person can state their own strength of opinion but cannot dictate how others vote or try and railroad this. This section is now getting unwieldy due to discussion here. I will either collapse, footnote or move more general discussion. Casliber (talk · contribs) 21:06, 19 May 2011 (UTC)

• Um, my only goal here was to urge people to be more focused and careful than normal, given all the talk of trouts and bans. I can't tell anyone what they should or shouldn't discuss. I've probably said everything I wanted to say at this point. - Dank (push to talk) 22:12, 19 May 2011 (UTC)

Comment:I totally prefer, against some apparently higher minds but in line with common usage, to use a hyphen with verses and a dash with "to stand for and between independent elements" . There has to be some common sense here that also goes along with compromise. If there is not a run to do away with the hyphen, that I am so totally against, and a move to use "Wiki" mark-ups regardless of common usage, then give meaningful reasoning for different usage.

• Also, PLEASE!; if someone is making comments, putting in "their two cents", or even giving civil comments, lets not use the "we have to do this now because of ArbCom and certain sanctions that are sure to come to involved editors". If I am civil and trying to help better Wikipedia and someone decides that for that reason I deserve sanctions then I will decide I have a far better use of my time than here. Apparently one editor is contemplating this. I just recently became involved and it took a while to attempt to try to catch up. I feel something that is deadlocked needs attention and an Arbcom is one way but threats just for the sake of it, and using this a lot as an argument to possibly force compromise, as apposed to compromise for the betterment of Wikipedia, is against what I perceive as a fundamental Wikipedia process. As stated above, "guidelines" for consistency is a good thing and needed to a point. If a majority of common usage (and not some 5% I observed) is one way then Wikipedia should lean in that direction. If a project is shown to use one certain way because common usage, consensus on the project, and other like titled articles, then there should not be a mandate to "do it wrong" just to be politically correct. If a title is otherwise correct (but just not within some views) the difference can be noted at the beginning of the lead. My goal here is the betterment of Wikipedia, threats of bans and sanctions notwithstanding. Otr500 (talk) 13:05, 20 May 2011 (UTC)

In compounds whose elements themselves contain hyphens or spaces

• Therefore we disagree. This provision does not have unanimity. Septentrionalis PMAnderson 20:26, 19 May 2011 (UTC)

Here again Anderson is claiming veto power over any provisions of the MOS. Whatever he does not agree with personally is therefore inappropriate. — kwami (talk) 02:48, 20 May 2011 (UTC)

Is he really, or did I miss something? His previous post is trivially true; it only says we disagree, therefore no unanimity. He isn't vetoing a provision of the MoS. The title of this section is "What do we agree upon?", not "What stays in the MoS?" Thus it is both relevant and undeniable than Septentrionalis does not agree upon this point. Art LaPella (talk) 03:43, 20 May 2011 (UTC)

• Comment; Are we requiring unanimity as a basis of compromise. And this provides solutions how? It seemed to me there was more "agree" than "disagree" so I go with Agree. Does this help solve the problem are are we just spinning wheels?
  • No, we are not requiring unanimity as a basis of compromise. Please review the top of this page: "Editors have from now until 30 May 2011 to finalise the questions on this page, and a further six weeks to vote on them. Editors are not prohibited from voting beforehand but strongly urged to doublecheck after May 30 to view the final structure." Unanimity is required to "finalise the questions" under this #What do we agree upon? header, because where there is unanimity, there is no need for a question about whether to change the guideline. Unanimity or disagreement are the only possible answers to "What do we agree upon?" During the further six weeks period, unanimity will presumably not be required for a decision. Art LaPella (talk) 14:42, 20 May 2011 (UTC)

I guess I am still missing something. If there is an agreement of three editors and I disagree this is consensus so why risk missing a "deadline" by requiring complete agreement. Is this just to get it on the shelf to avoid the deadline and end up with further deadlock on an attempt to find consensus at that time? Although everyone getting along and world peace is a grand goal for Miss America I question unanimity in this situation. I am just stating my opinion and feel that if I am in the minority of a fair discussion my comments should be noted and over-ridden. I was to the opinion that that was a key part of the Wikipedia process and certainly could be helpful here to 1)- Avoid stretching things out and getting close to the deadline with no solutions, and 2)- Avoid extending the issue to a later date when not "under the gun". I do understand (I think) the reasoning that if something is brought up and all that comment agree then it is settled. Of course there are editors that have been active in the debates that has either refrained from comments (maybe for fear of sanctions) or have just not yet weighed in. At any rate until the button is pushed that states this phase is over, and until that time one person could stall the process that might otherwise seem to have unanimity, everything requiring the purported unanimity could be stalled. I just do not want people to state an agreement, to get past a deadline (and feared possible sanctions), that they will vehemently oppose at a later date and create problems then. I think my assumption is correct that only listing what we all agree on will presumably simplify the process of getting on with the next phase but why (a question to help me understand) require complete agreement when this is not even the acceptable Wikipedia process? Yes this may get us over the "what we all agree on" parts but they were probably not the real issues to begin with. I can not imagine that forcing a false agreement was the goal of Arbcom. These are just my thoughts and not meant to hinder any process that might work. I will try to weigh in lightly with the mandate in mind but certainly seeking improvement to Wikipedia as the major goal. Otr500 (talk) 16:04, 20 May 2011 (UTC)

OK, I'll try again. The arbitrators have asked us to separate the process into stages. This stage is simply to list disagreements (or to list agreements in order to exclude them from the disagreements). This stage is not to decide which side of a disagreement will prevail. The only way to list disagreements is to see who disagrees. The only way to see who disagrees is to let anyone disagree. If anyone disagrees with an issue, then that issue should be one of the disagreements. Adding to the list of disagreements shouldn't risk missing a deadline. Adding to the list of disagreements doesn't mean that the disagreement should prevail. Adding to the list of disagreements doesn't "stall the process that might otherwise seem to have unanimity". It only adds to the list of disagreements. Only in a later stage can we see who else agrees or disagrees, and determine a consensus. The arbitrators' plan is first we "finalise the questions" by determining where we disagree. Only then do we answer those questions. You haven't criticized the arbitrators' plan, so yes you are still missing something. Art LaPella (talk) 18:31, 20 May 2011 (UTC)

But already, we have partial disagreement on almost everything but item 1 (and perhaps 4/6), so unless we separate the distinct uses to the point where the only possible opinions are “agree” or “disagree”, we’re not going to find much unanimity. And if a hyphenist decides to join the discussion, we may not even find unanimity on any point. Finding unanimity may be tough even with everything broken out, but until we try, we’ll never know. It would seem to me that a breakout would almost require a complete re-vote (save perhaps item 1), so the longer we wait to do this, the more effort gets wasted. In the case of item 4, I’m still not even sure what that usage is, so I suppose I can’t really say whether I agree or disagree. Again, I think we may have skipped a key initial step: identify the uses on which we need to decide whether we agree or disagree; the sooner we address this, the better. JeffConrad (talk) 20:37, 20 May 2011 (UTC)

I should add that the breakout I′ve suggested derives mainly from the examples in the current MOS, and may not be sufficient. For example, what do we do with a batter who is 1–3 (one for three) and facing a 3–2 count? And I’m sure there are many others. The style guides are deficient here; for years, I wasn’t sure whether Chicago supported the en dash in the sense of versus or between; it was not until the specific disclaimer in the 16th ed. that I learned they did not (save perhaps situations covered by Scientific Format and Style). Some things still aren’t clear: Chicago suggest an en dash to distinguish university campuses (University of Wisconsin−Madison); by this logic we would have Herb Kohl (D–Wis.), but I hardly ever see the latter construction. I’ll concede that, unlike WP, most style guides don′t have infinite space to devote to this topic. JeffConrad (talk) 21:50, 20 May 2011 (UTC)

On these matters, silence implies dissent. The general rule is to use a hyphen for compounds, even in the wilds of Hart's New Rules. where a dash is not expressly preferred, a hyphen is preferred. As for your example, the batter should be 1 for 3, and facing a 3-2 count, especially if the writer uses a dash for ranges. The same symbol should not mean different things when there is a distinct symbol to use. Septentrionalis PMAnderson 22:54, 20 May 2011 (UTC)

That’s your take, but not necessarily everyone’s. As an example, the APA style guide doesn’t mention en dashes in ranges of numbers, yet they use such dashes themselves. The MLA guide doesn’t even mention en dashes, yet they also use them in ranges of numbers. So I don’t buy “silence implies dissent”—silence may simply mean they never thought of it.

On a couple of specifics: we agree on “1 for 3”, but not necessarily on “3-2 count”, which no one who knows anything about baseball would confuse with a range of numbers. I would use the latter as I showed it, but might be persuaded by overwhelming usage to the contrary, e.g., Herb Kohl (D-Wis.). JeffConrad (talk) 23:43, 20 May 2011 (UTC)

In compounds whose elements contain hyphens—moved from poll page

5a. In compounds whose elements themselves contain hyphens (the anti-conscription–pro-conscription debate).

• Agree (somewhat reluctantly) This construction strikes me as ugly, but in some cases may be the best option. I would suggest that editors consider recasting when feasible. JeffConrad (talk) 08:36, 4 June 2011 (UTC)

[other comments elided]

• Disagree. This usage (with the next one below) is as exclusively American as they get; and it is not universal in American anyway. It generally makes sense only on a background assumption that the en dash with sense "between" or "versus" is not allowed. That is why it belongs in CMOS, for example, but not in our MOS. In the current example the en dash is already justified on those other grounds. If we would write the radical–conservative debate, we should write the anti-conscription–pro-conscription debate, hyphens or no hyphens. I struggle to find plausible examples that work differently. "A Mon-Khmer–native-speaking woman"? "A long-established–anti-intellectualism problem", to distinguish from "A long-established anti-intellectualism problem"? [Added later: The en dash introduces a new possible reading, so the ambiguity is still as great: "A long-established versus or and anti-intellectualism problem".] "Little-known–anti-inflammatory contraindications", to distinguish from "Little-known anti-inflammatory contraindications"? This guideline was added late, and was not well considered or well drafted. It fits ill with the more settled guidelines. Noetica^Tea? 00:28, 27 June 2011 (UTC)

  I could as reasonably claim that the spaced en dash is almost exclusively British. I’d be careful with the royal we . . . I realize CMOS isn’t your favorite guide, but it probably handily outsells every British guide combined; that doesn’t make it the standard, of course, but certainly argues against its being completely blown off. And the usage isn’t strictly CMOS anyway. We of course need to be careful to be clear on what we mean by “usage”. None of the guides, including CMOS, seem fond of using en dashes with hyphenated compounds, so I have no issue with discouraging them here. But I take issue with insisting on post-Civil-War period over post–Civil War period, or New Zealand-style clothing or New-Zealand-style clothing over New Zealand–style clothing, which I read OSM as preferring, even if it’s only as the least of evils. We could of course chose to commemorate Fowler with NewZealand-style clothing. JeffConrad (talk) 01:33, 27 June 2011 (UTC)

  Jeff:

  1. Do you mean the spaced en dash in ranges, or what? Much publishing in America uses spaced en dash rather than em dash at the sentence level.

  2. What "royal we" are you talking about? Here I use "we", as a normal expressive device, only in "If we would write ..." and the rest of my conditional statement. It isn't about "us", it's about anyone who wants rational and consistent punctuation.

  3. I make no point here against CMOS; I use CMOS myself. Nor do I say that this guideline is unique to CMOS. I only say that this particular guideline fits well with CMOS (which appears to have invented it, a century ago). And I demonstrate that it fits ill with the other guidelines in WP:MOS, some of which make it practically redundant: I show that the result will often be the same, but from differing principles.

  4. The topic here is not cases like post–Civil War period. They involve spaces in a component, not hyphens. Let's keep points in their appointed places.

  Noetica^Tea? 02:12, 27 June 2011 (UTC)

  I somewhat misspoke. I clearly reacted to “hyphens or no hyphens”, the latter not covered by this section. I also reacted to “That is why it belongs in CMOS, for example, but not in our MOS”, perhaps overly so—but it’s hard to take this as other than a ding, however slight. So to an extent we both misspoke. As for “we”? Perhaps I incorrectly read you as claiming to speak for everyone here, especially as strongly as you did. Though I’m not a fan of the combination of hyphens and en dashes (it’s sometimes tough to quickly distinguish one from the other), I’d stop short of implying that it’s irrational. Perhaps more important: CMOS aren’t fans of it, either, as their recasting illustrates. My greatest issue with the examples in any of the recent editions of CMOS is that they most likely intended “quasi-legislative, quasi-judicial body”, a difference irrelevant to the issue at hand. As for spaced en dashes in the US? They’re very uncommon in any application, and find almost no support in American style guides. And OUP agree with me . . . but perhaps we’re both wrong. In any event, we probably should be careful with tags such as “exclusively American” or “exclusively British”, because neither is better than the other; you won’t find me writing speciality, but I don’t expect you to write specialty, either (though you may note that I usually treat collective nouns as in BrE). Finally, we really don’t differ that much on combining hyphens and en dashes in phrasal adjectives, though I would discourage it rather than ban it. Several others here seem to agree. JeffConrad (talk) 02:59, 27 June 2011 (UTC)

  We should probably move most of this to the discussion page. JeffConrad (talk) 03:11, 27 June 2011 (UTC)

  Jeff, I am happy to leave it where it is, provided there is no more to say. There's no point of yours that I need to come back on. I'm content that my points (raised by no one else above) have been noted. Applying [small][/small] to our exchange. Noetica^Tea? 03:46, 27 June 2011 (UTC)

[moved to end to maintain chronological order]

“hyphens or no hyphens”: The latter isn’t covered by this section. JeffConrad (talk) 02:59, 28 June 2011 (UTC)

I don't understand why you repeat that, Jeff (see below above). The point is that the example would already have an en dash without those hyphens. Imagine if we were to spell the components solid: the anticonscription–proconscription debate. And compare this, which is almost synonymous: the volunteerism–conscription debate. In all three cases a long-standing principle in WP:MOS (separately examined on this page) already calls for the en dash. So the principle that we focus on right here is not shown, by its current example, to have force of its own – unless that other principle is removed. We can't deal with each principle in hermetic (nor eremitic) isolation. By all means start a new section at the other subpage, removing our exchange to there, if this is not yet clear. Noetica^Tea? 04:04, 28 June 2011 (UTC)

I put it here because it was the main reason for my earlier comment; I assume that put in small type, the rest of our exchange will be ignored by most readers. If all we really are discussing here is the combination of hyphens and en dashes in a string without whitespace, we have no disagreement. And CMOS essentially agree with us. And I don’t see how we got into eliminating the hyphens (unless you took NewZealand seriously; even then, the caps might suffice). If you suggest we disallow en dashes in open compounds because they’re “exclusively American”, we do disagree. But if so, this is not the place to discuss it. JeffConrad (talk) 04:49, 28 June 2011 (UTC)

Jeff: First, refactor our exchange as you will; or remove it all to the other subpage. I don't care; and I certainly trust you to do it well. Second, CMOS (for example! I have no special point to make about or against it here) wants a hyphen in the volunteerism~conscription debate; so CMOS and WP:MOS differ on that. This is directly relevant to what we discuss here, as I have explained. Because CMOS wants that one to have a hyphen, it needs a special, extra principle to get an en dash into the anti-conscription–pro-conscription debate. Agreed? So my point (to repeat it again) is that WP:MOS does not need such an extra principle, since it already calls for an en dash in, say, the volunteerism–conscription debate. That same principle would give us the anti-conscription–pro-conscription debate also. CommasHyphens in the components make no difference.[Fixed a slip in that sentence. See discussion below.–Noetica] This was highly relevant to raise here, even if it implicates another of the principles – one that we discuss elsewhere on this page. Once again: it is reductive and unproductive to insist on examining each separate principle in strict isolation. I oppose this present principle, largely because another principle in WP:MOS (one that I support, with the majority here) makes it redundant. If you disagree, please show a plausible example where this CMOS-style principle would apply under WP:MOS, and where the WP:MOS "between" en dash would not already be called for (that is, before we even consider whether or not the components have hyphens). As you can see, I tried; but I could not find any such plausible or clarifying application of this principle. Noetica^Tea? 05:30, 28 June 2011 (UTC)

[New comment after move]

I agree that CMOS would use a hyphen in the volunteerism~conscription debate (specifically, 6.81 in 16th ed.), and that usage differs from our MOS and from what I’ve voted for here. But I don’t see how this enters into this specific issue. I agree that WP:MOS arrives at the anti-conscription–pro-conscription debate, but I am not sure of your position—do you disagree with the usage or simply that it is a special case? If the latter, I guess I’m to blame once again for breaking this out. I did not do so capriciously, however. Both CMOS and OSM deprecate this usage, as do I. Though the anti-conscription–pro-conscription debate is better than the anti-conscription-pro-conscription debate, many readers may not see any difference, which I suspect is why CMOS and OSM dislike it. And my breakout was not intended to suggest how the MOS might be rewritten—I was simply trying to find where we had reasonable agreement and where we did not.

I’m also puzzled by the reference to commas just above; the only example I can find in CMOS16 that includes a comma is a quasi-public, quasi judicial body (at 6.80), which is how I would write it, and is a construction I would read our MOS as endorsing.

Again, I wonder if we seem to be disagreeing on something on which we really don’t disagree. JeffConrad (talk) 07:14, 28 June 2011 (UTC)

Jeff, thanks for moving it all to here. Too long for the voting page. And I'm sorry for my mistaken use of the word "commas" when I meant "hyphens" (now struck out and corrected above).

Look, it's all very rational; but since short explanations have not worked, I must try a long one:

I oppose including a principle like CMOS's one, especially with WP:MOS's present example: the anti-conscription–pro-conscription debate. Why? Start with this fact: that example could equally have been used in MOS (as it now stands) to illustrate the ordinary use of a "between" en dash. Look at it (without "the"), among the other examples on the voting page for the "between" dash:

male–female relations

French–German border

anti-conscription–pro-conscription debate

For each, the abstract schema is this:

Schema 1: A–B C, where "–" represents a relation similar to "between", and the compound A–B modifies C

Right? Now, in two cases the components are simple. Neither includes a hyphen. In the third case, the components include hyphens. But nothing in the schema, or in the guideline as it stands, precludes that. [NOTE: In this explanation I focus on "between"; that is perhaps the meaning most relevant in the example anti-conscription–pro-conscription debate, and it is the one explicitly excluded in CMOS. But I do mean to include "versus", "to", etc. in Schema 1. Focusing on "between" makes the explanation easier to write and to follow.]

Given that our Schema 1 for the "because" cases covers anti-conscription–pro-conscription debate, any separate principle that uses it as an example – trading on the presence of hyphens, now – is redundant at best and misleading at worst. This is not a problem for CMOS: we agree, it does not support the "between" use of en dashes; yet it does want anti-conscription–pro-conscription debate. Several guides agree with CMOS, and make similar provisions; MOS is among the many that do not agree with CMOS, and it would be unhelpful for us to adopt a principle from it without careful thought.

Of course I agree with the phrase anti-conscription–pro-conscription debate! Where did I suggest that I don't? I have explained: my support of the "between" en dash entails that I support anti-conscription–pro-conscription debate, just as it entails that I support volunteerism–conscription debate. I can imagine that the added CMOS-based principle would be useful in other cases, with this schema:

Schema 2: D~E F, where "~" marks D as modifying E, and the compound D~E itself modifies F.

In such cases, when D and E are simple, CMOS and WP:MOS agree that "~" becomes a hyphen (generally; CMOS16 makes some changes):

small-state senators (in CMOS16, 7.85, p. 377)

But by the CMOS-based rule, if D and E each included a hyphen and D still modified E according to Schema 2, "~" would be realised as an en dash, not a hyphen. Something like this, breaking D and E into smaller parts:

Schema 2': d'-d"–e'-e" F, where d'-d" modifies e'-e", and d'-d"–e'-e" modifies F

I seriously doubt that CMOS ever envisaged such a thing; but technically CMOS16 6.80 endorses it (always with the preference for rewording, granted). Not a problem in practice, because for CMOS their 6.80 rule does something useful: it allows for the more normal cases in which the components have hyphens, like anti-conscription–pro-conscription debate, while a parallel case with simple components has a hyphen, not an en dash: volunteerism-conscription debate. That's what they're after.

Now, accept the onus of proof. Show that we have any use for that CMOS-based principle. Since you agree with the "between" en dash (as I see from the voting page), what use do you see for this added CMOS-based principle? This is not politics: it's not about the relative merits of different style guides, or British–American differences. Just tell me what purpose this added principle would serve. MOS already provides for Schema 1 cases with an explicit "between" principle. So for MOS, the only cases in which the CMOS-based rule could make any difference are the Schema 2 cases. Can you think of such a case where an en dash would be at all plausible, and not misleading? So far I cannot! See above, where I attempted these:

a Mon-Khmer–native-speaking woman

a long-established–anti-intellectualism problem

little-known–anti-inflammatory contraindications

I explain above how these are strange, misleading, or both; but they are the only sorts of phrases I can imagine the CMOS-based principle specifically permitting in a WP:MOS context, where anti-conscription–pro-conscription debate (for example) is already permitted by the plain "between" principle. Show me different.

Noetica^Tea? 10:04, 28 June 2011 (UTC)

It appears that your original answer was meant to the effect of

• I agree with the usage illustrated by anti-conscription–pro-conscription debate, but disagree that it is a separate case, because it’s already covered by the sense of between. CMOS and OSM list it as a separate rule because Chicago do not use the en dash for the sense of between, and because both CMOS and OSM find the construction awkward.

Had you written it as such, I might have said nothing. After several re-reads, it looks as though this is kinda what you did say, but compare with what I wrote just above (I concede I may have been overly distracted by exclusively American and CMOS).

Again, again, again, the separation I made was to see where we stood on various usages, not to vote on current or potentially new wording of the MOS, which would be getting ahead of ourselves by presupposing the outcome of the polling. It was necessary to break out 5a because some, including CMOS and OSM, deprecate this usage even though they allow it. There clearly are some here who agree; were we to deprecate this usage (urging recasting when practical), it would need to be a separate case—call it an exception to the general rule for the sense of between if you will. I think you extrapolate Chicago’s rule far beyond anything they would ever suggest. They might reluctantly accept anti-conscription–pro-conscription debate as preferable to anti-conscription-pro-conscription debate, but prefer recasting because they recognize that most people may see the former as if it were the latter anyway.

So again, it seems that we really don’t disagree that much, except that I would recast this usage where practical.

As for the next usage (compounds with spaces), I disagree that it is “as exclusively American as they get”, because OSM allows it as well, even if OUP like it less than do Chicago. A little definitely goes a long way, but I simply cannot see Chuck Berry-style lyrics as a substitute for Chuck Berry–style lyrics. Some here obviously disagree, just as there are some here who would ban the spaced en dash as an alternative to an em dash. The former is just about as British as Chuck Berry–style lyrics is American. If it′s not about politics, neither should matter. JeffConrad (talk) 21:28, 28 June 2011 (UTC)

Jeff, by usage in my original statement I meant all that the stated principle would allow, not the example given. And I then wrote this, including the bold that I show here: "In the current example the en dash is already justified on those other grounds. If we would write the radical–conservative debate, we should write the anti-conscription–pro-conscription debate, hyphens or no hyphens." It is not always possible to predict how one will be read, or how carefully. It is unfortunate that we failed in sending and receiving the message this time.

Yes, OSM makes almost the leanest possible concession to the usage we discuss here, and to the next. Its successor NHR (typically less expansive than OSM) allows less; and it specifically labels that next usage as American:

Note that in US style an en rule is used to connect a prefix and a compound (the post–World War I period). (NHR, p. 55)

I stand by my initial statement: "This usage (with the next one below) is as exclusively American as they get; and it is not universal in American anyway." I did not say, without qualification, that it was exclusively American!

Now, again (and again) I ask you: in a set of WP:MOS guidelines that includes the "between" use of en dash (and the "versus", and the "to", and the "and", as at present), is there a role for this special CMOS-based rule? Do you see that it could only be needed in this sort of case:

d'-d"–e'-e" F, where d'-d" modifies e'-e", and d'-d"–e'-e" modifies F

As far as I can see, all other cases are already covered by the "between" principle (and those others for "versus", etc.). If you do see a place for this CMOS-based rule (a neutral descriptive term), please show examples where it would be useful in increasing clarity. So far I genuinely, apolitically, cannot find any. (I haven't had time to search thoroughly.)

(I have not yet voted on those cases involving spaces as opposed to hyphens, so I do not address them above except to comment on NHR, etc. More on them later.)

Noetica^Tea? 00:59, 29 June 2011 (UTC)

How many times must I say it? I have no disagreement that compounds that contain hyphens are already covered by several of the senses that would call for an en dash, provided we agree on those other rules. Seems to me the purpose of the poll is to find this out, so we cannot assume the answer while posing the question. The responses suggest that many editors, while they would allow the usage, they would sometimes encourage recasting, and they would have strong support from CMOS and OSM. This may argue for a special rule or an exception to the more general rule—I’m not sure it really matters what we call it. Let’s see what we agree on before we act on what we agree on.

I recognize that CMOS and OSM arrive at the special rules here for different reasons; for CMOS, it’s a special case of usage, and for OSM, it’s an exception to the general rule. But the results are similar.

I think “exclusively American” and “exclusively British” (perhaps with a few qualifiers) serve only to distract, and encourage endless unproductive discussion. I could cite OSM as saying that the spaced em dash is as exclusively British as they get:

OUP and most US publishers use the unspaced (non-touching) em rule as a parenthetical dash; other British publishers use the en rule with space either side. (at 5.10.10)

It should be obvious that most AmE editors favor American practice and BrE editors favor British practice; it would surprise were it otherwise. As long as a practice is well established, it should not matter where. What we want to avoid are fringe usages with no rational bases. It probably could be credibly argued that en dashes are a fringe usage anyway, and that most readers would not know an en dash if it bit them. Though I’ve been part of that fringe for years, I find our extended interest in en dashes amazing when we reject nearly universal convention for quotation marks (noting that even here American and British practice differs)—most folks deal with the common at the expense of the less common. But I don’t want to go at that one here—we can create a subpage dedicated to that topic if necessary.

Once again, I agree that the specific usage that this question addresses is covered by the more general rule, but if it is decided that we, like Chicago and OUP, prefer recasting of some constructions, we need to give this usage special treatment. We can decide later whether it’s an exception or a different rule. JeffConrad (talk) 02:54, 29 June 2011 (UTC)

Jeff: You may protest that you have to say things many times; but still you do not answer my repeated direct questions. Therefore I will do the necessary work myself. A hypothesis, just so I can be absolutely sure of something I consider important:

JeffConrad and Noetica agree on this: If the existing WP:MOS principles for en dash indicating "between", "versus", "to" and the like are retained, there can be no justification for a further principle involving components that include hyphens. Such cases are already completely covered by those former principles. The only exceptions would be rare, implausible, and very likely misleading if an en dash were introduced.

Please answer: Is my hypothesis right? Do we agree? Please state clearly and completely how we disagree (preferably with examples). Alternatively, please say why we need to talk about anything more here, and say what that unfinished business might be.

Your remark about discussion of British and American English being a distraction, and so on? I read it as distracting! Why raise all that here? We have moved beyond any British–American divide, I had thought: I now want to know how you envisage the guidelines we favour for WP:MOS working together, apart from any considerations of other style guides, or of varieties of English.

To answer your own recent point above: I do not touch on the business of recasting, because if the example given – the anti-conscription–pro-conscription debate), along with others like it – is indeed covered by the former principles ("between" en dash, etc.), advice to recast belongs in the presentation of those principles. For the specific instances in which hyphens are present, right? It would be strange and artificial to add a redundant separate principle (this CMOS-based principle) simply to state that it will not apply, and that recasting is desirable despite what the former principles seemed to call for. Understood?

Noetica^Tea? 03:31, 29 June 2011 (UTC)

As I’ve said several times, if your hypothesis is that the more general case covers this usage, then, yes, I agree with you. The question asked in this poll, again, was with the specific usage—nothing else. Had your comment been “I agree with the usage illustrated by anti-conscription–pro-conscription debate, but disagree that it is a separate case, because it’s already covered by more general rules”, I’d probably have said nothing.

I wasn’t the one to open the American/British issue, and see no point in continuing with it if we both think it’s largely irrelevant. I would not necessarily deprecate a practice that was primarily one or the other, possibly because we might never find consensus on which practices are “as exclusive as they get”.

At this point, I don’t care where an issue such as recasting might be addressed—we’re polling on usage, not rewriting the MOS. But if the recasting arises solely because the combination of hyphens and en dashes seems awkward, then recasting advice must relate specifically to that usage, as is done in both CMOS and OSM. How and where it’s done here is something to address if we decide to change the MOS.

Finally, it seems as if we’re spending an awful lot of time parsing minor details of something on which we largely agree. JeffConrad (talk) 04:55, 29 June 2011 (UTC)

Rightio, one circuitous way or another I think we've covered everything on this narrow topic. Finally (I hope), it would be useful to record and acknowledge that no one, in the reams of dialogue on these pages, had pointed out what I point out in my vote and in this subsequent discussion. It has a definite bearing on how amended guidelines ought to be cast – though obviously we have disagreed about where it is appropriate to announce such a momentous discovery. Again: ☺. Noetica^Tea? 05:15, 29 June 2011 (UTC)

I think it still depends on what we decide we agree on. If, for example, we decide that we agree on the usages covered by the current (2) and (3), and agree that “Unspaced en dashes are used in place of a hyphen in compound modifiers (see under Hyphens above) in which the parts of the compound are independent and equal elements”, adequately describes these usages, and we decide that recasting may be preferable in some instances, the treatment of compounds that contain spaces or hyphens could be treated as an exception to the general rule. If, however, we decide to retain more than one general rule, it would seem better to have a rule that dealt with hyphens, spaces, or both than to have exceptions repeated in two different general rules. At this point, I’m not sure that “independent and equal elements” would obviously include ex–prime minister or Chuck Berry–style lyrics, so I think dealing with the wording before tabulating and interpreting the results is putting the cart ahead of the horse. And I’m not sure continuing to beat the horse until its morale improves is the best approach. Let’s see what we agree on, see what rules that suggests, and if revision of the MOS is indicated, deal with the details at that time. JeffConrad (talk) 09:41, 29 June 2011 (UTC)

When prefixing an element containing a space—moved from poll page

5b. When prefixing an element containing a space (pre–World War II technologies, ex–prime minister).

• • Disagree. Like the preceding principle, this one is almost exclusively American. By itself, that is not an important objection; but these two principles originate and belong in a more limited context than English Wikipedia, which serves the whole anglophone world. It may fit well with CMOS conventions, for example. But our other guidelines will surely continue to differ from CMOS on en dashes (and dashes generally). Moreover, when we consider cases of this sort in isolation, we find that the structure is usually clear with the standard hyphen anyway: from the capitalisation (pre-World War II technologies), or from the inclusion of more hyphens when the prefixed form is not a proper name (post + blue period + paintings: post-blue-period paintings), or from the use of a stock title or familiar pattern that is read as a coherent block (several ex-prime ministers were present; two anti-potassium cyanide agents, noting that such chemical names do not take hyphens). In rare cases there may still be uncertainty of meaning: are non-sodium chloride salts chloride salts that are not sodium chloride, or simply salts that are not sodium chloride? Such cases are rare, and can be resolved by context or by rewording. In conclusion, if we do adopt this principle, we do so for extremely few cases, against expectations for many users of the language, and with juxtapositions that are, as we say in the trade, butt~ugly (post–French Baroque pre-Revolution art), and without a clear way to extend to more complex constructions by other, more established principles (quasi-pre–French Revolution taxes??; quasi–pre–French Revolution taxes??). Noetica^Tea? 05:04, 12 July 2011 (UTC)

    I would again caution on the use of the royal we; although some obviously share your views, others obviously do not. I would also remind that, when referring to anglophones, one should be mindful of the portion thereof that Americans constitute. Now as people rush to throw in India and whomever else the Brits invaded before the US acquired the hobby, I’d like to see specific citations of usage rather than vague allegations.

    Is post-Cold War period perfectly clear? Perhaps, and the capitalization definitely helps; but should we have a different rule when there is no capitalization? And even with the capitalization, the reader must backtrack, if ever so slightly. As I’ve indicated several times, it’s not a matter of the reader not being able to figure it out—in most cases, the reader can easily do so. But the reader should not require any additional effort to do so. If it is otherwise, nearly all other instances of using the en dash are unjustifiable.

    Perhaps a clue can be taken from NHR, which concedes (at 3.3.4), that there is no satisfactory way of dealing with ex-Prime Minister. In that specific instance, recasting to former Prime Minister may work, but the same cannot be said for pseudo-page transition. Because NHR indicates that US practice at least has a solution, it would seem sensible to at least allow it if the objective is truly “best practice” rather than British practice.

    Be assured that I think this approach has limits; though I’m fine with the first two examples in CMOS16 at 6.80, I’m considerably less enthusiastic about the third, and utterly unenamored of the last. JeffConrad (talk) 09:06, 12 July 2011 (UTC)

    Thanks for moving things to here, Jeff. See my brief response at the voting page, and this here:

    I am confident that you do not mean any offence; but in fact none of those is a "royal we"; nor was there any on the previous occasion that you imputed such a usage to me. The nearest candidates this time: "Moreover, when we consider cases of this sort in isolation, we find ...". This is a perfectly standard usage in a register of English I am very familiar with, and it is widely understood. It is a way of expressing a point that one judges well evidenced, not a mere statement of idiosyncratic opinion, with an appeal to the reader to join in considering and accepting that evidence. To express something more idiosyncratic, I should say (and do say) "for myself", "it seems to me", or "I suppose". It would be a "royal we" only if I referred to myself alone as we. I caution you in turn to respect the expressive style of editors who post here, no matter what style they choose or are limited to. It is an element of civility on Wikipedia to do that. Getting the message across clearly, efficiently, and unoffensively is good enough.

    We could go on at great length about the content here, and what you take exception to. I propose that we not do so. The voting seems slightly in favour of the principle in question (note the hedging in several agrees though); so I have decided that I will not stand against it being retained in the guidelines, unless there is a last minute rush to exclude it. Many agree that re-wording is often better. I have simply recorded some observations from my own rather extensive investigations and experience; you too have extensive expertise, and of course I respect that. There is no practical consequence of any disagreement we might discover between us right here; so let's leave it. Noetica^Tea? 09:58, 12 July 2011 (UTC)

    Be assured that no offense was meant. Perhaps we are simply afflicted with the division of a common language . . . at least in AmE, it’s difficult to take we as other than presuming to speak for a significant majority, seemingly confirmed by “well evidenced” above; perhaps it’s different in BrE, and perhaps I need to learn this. We could probably debate this forever, and most likely would quickly realize that what’s “well evidenced” to some is not evidenced at all to others. Ultimately, I think we must recognize that there’s more than one way to do it.

    Perhaps an expansion of my position (and what I think CMOS intends) is in order. What we essentially have is an issue of operator precedence. Clearly, we all seem to agree that a hyphen binds more tightly than a space; in this instance, there is an attempt to use the en dash as an operator with lower precedence than the hyphen or space, so that Chuck Berry–style lyrics is read as “(Chuck Berry)-style lyrics” rather than “Chuck (Berry-style) lyrics”.

    Again, recognize that I only support this usage up to a point; I would probably recast country–music influenced lyrics to the alternative that Chicago give. But there is no way I would go with pseudo-page transition over pseudo–page transition. Though I’ve given this as somewhat of a running joke, I assure you it was no joke when I first encountered it 25 years ago, and if I am to judge by every book on troff that I’ve ever read, I wasn’t the only one baffled by it. Would use of an en dash as Chicago recommend have changed the world? We’ll never know, and it no longer matters. As with so many other uses of the en dash (on most of which, you and I agree), this fine distinction will probably be lost on the majority of readers. I would say here what I’ve said for the other uses: though many may not notice, those who do will be aided, and at worst, it does no harm. JeffConrad (talk) 10:45, 12 July 2011 (UTC)

Spacing of endashes

The space before an en dash should preferably be a non-breaking space ( ).

• Not sure. — kwami (talk)
• Don't get me started. - Dank (push to talk) 14:58, 19 May 2011 (UTC)

  ...even because that would then get me started. ― A. di M.​_plé​^dréachtaí 17:12, 19 May 2011 (UTC)

• I'm not sure how to interpret Dank's comment, but this isn't controversial, is it? This just means you don't want a dash/hyphen/whatever at the beginning of the line, right? ErikHaugen (talk | contribs) 16:55, 19 May 2011 (UTC)
  • I'd just rather not get into it right now when sanctions have been promised if we don't finish in time. I haven't seen anyone demanding resolution on this as part of this Arbcom case. - Dank (push to talk) 17:35, 19 May 2011 (UTC)

    Ok, so just leave as is (for now)? ErikHaugen (talk | contribs) 17:54, 19 May 2011 (UTC)

    Agreed, leave it alone. - Dank (push to talk) 18:24, 19 May 2011 (UTC)

    I tend to make the same interpretation of the ArbCom motion as Dank, but fail to see how anyone could be “sanctioned” for failing to reach consensus. What exactly would a sanction be? JeffConrad (talk) 20:55, 19 May 2011 (UTC)

• Agree, as Erik has stated. Too much trouble? Use {{ndash}} or use an unspaced em dash. JeffConrad (talk) 19:50, 19 May 2011 (UTC)
• Comment; I agree with JeffConrad and the continual use of the "threats of sanctions" if consensus is not met is not how the ruling appeared to me. I can see that the ordered part of admins being proactive is not being followed or these continued mentions of such "threats" would be a good time for clarification. I guess some that worry about this so much as Wikipedia is their lives, with nothing else to do, this might be a motivation for compromise or a time for silence. I also feel that compromise should be a two-way (would that be with a hyphen or en-dash) street (to agree) with Wikipedia interests, reader interests, and other editor's interests, in mind. I think the "sure to come" sanctions are concerning civility and disruptions. At least I would hope this is the intent of the rule (it is somewhat— or maybe a lot "vague") as I feel that would be an over-reach of authority. Again as stated, if I am sanctioned or banned for trying to improve Wikipedia then that would be a big hint I need to find better use of my time. I have noticed that some editors that have been very active has not made comments and feel that if this is a serious issue any editor can seek clarification at the comment section of the Arbcom ruling. With that in mind (ask if you are worried) can we be more concerned with improving Wikipedia than some sanctions that we are not even sure about. Otr500 (talk) 13:05, 20 May 2011 (UTC)

  The ArbCom motions concludes with “If after two months, a determination isn't realised, a case will be opened and conduct violations will be dealt with severely.” I think fear of sanctions may be reading too much into this, though clarification of “case” and “conduct violations” would be helpful. Looking practically at the two-month deadline: I can’t imagine anyone able to stomach this discussion for two months (OK, perhaps a few . . . JeffConrad (talk) 20:22, 20 May 2011 (UTC)

  Look, I edit Wikipedia for my personal enjoyment. It's not a game, but... I mean, it's not "Wikipedia is my life", either. I feel that I can contribute, so I do. I think that I could contribute here as well, but I'm not going to get further involved in a topic that the most authoritative body on Wikipedia sees a need to "shepherd", and is making vague threats about the possible role they could play in the Wikipedia "careers" of those expressing their opinions on this matter. It doesn't matter what arbcom's intent may be here, the fact is that they're "watching", which is plenty of (de)motivation for myself (and apparently others) to just say "screw that, I'm not getting caught up in this". I mean, look, User:Casliber has already put up a big yellow warning message box at the top of this page, and added at least one authoritative "facilitator" message in order to steer the discussion. Screw it, it's just not worth getting further involved in this. I'm just waiting to see what you all decide, and I'll take what the MOS page says into consideration, now and once it's been changed.
  — V = IR (Talk • Contribs) 21:04, 20 May 2011 (UTC)

I note that some of the most frequent Manual of Style contributors haven't commented here. Maybe that's why, or maybe people are having trouble finding this page. For myself, if "a determination isn't realised", I'm confident I won't be accused of "conduct violations" unless I start personal attacks, or ignoring warnings or something. However, those participants who have a hard time avoiding personal attacks may prefer to stay away. Art LaPella (talk) 21:42, 20 May 2011 (UTC)

I'm fairly confident that I haven't done anything wrong either, but how do you know what arbitrators (or their enforcement squad, for that matter) will consider a "conduct violation"? That term means different things to different people, and different things at different times, after all. It's not at all clear to me that their limiting their oversight to obvious personal attacks or ignored warnings... like I said, look at the authority that has already been exercised on this page. I'm not paranoid (nobody is after me), but... I mean, come on, why participate here at all at this point?
— V = IR (Talk • Contribs) 22:24, 20 May 2011 (UTC)

• Agree. Leave it alone for the time being. [Could someone collapse the blowout above?]. Tony (talk) 04:08, 28 May 2011 (UTC)
• Agree. Unless we have specific proposals for changes, I'd leave it alone. Dicklyon (talk) 08:26, 30 May 2011 (UTC)

En dashes in article titles

• I can see where you are coming from, but the idea that it's silly to use “eye–hand span” in the text of an article titled “eye-hand span” or vice versa should be mentioned somehow somewhere. ― A. di M.​_plé​^dréachtaí 12:49, 19 May 2011 (UTC)

  When we get done with this exercise, I shall propose a text consisting of the points on which we agree. At that point we can enlarge it by consensus. Some wording about the desirability of an article spelling its subject like the title may well be consensus. (There may be exceptions, where the article uses a "correct" spelling and the title doesn't.) Septentrionalis PMAnderson 20:31, 19 May 2011 (UTC)

  If agreement takes “unanimity” literally, we seem to agree only on the current item 1, which I think we could have predicted. (and there are a few, such as those that insist on no dashes whatsoever, who probably disagree even on this). We might find agreement on a few others if we separate some of the different uses within the current categories, but even then I′m not sure we can find unanimous agreement. JeffConrad (talk) 20:50, 19 May 2011 (UTC)

En dashes in article titles_2

• • Wrong! Consistency across the encyclopedia is impractical, undesirable, and unimportant. On the style points on which English speakers most visibly diagree, such as the Oxford comma, the MOS already explicitly permits using either; our readers will - except for the most provincial ignoramuses - understand either. Septentrionalis PMAnderson 17:49, 20 May 2011 (UTC)

I am not sure how you got "Consistency across the encyclopedia" from a reply that included agreement with an editor that stated, "the title and the article should be consistent". If someone agrees with what someone else says there seems to be no sense in repeating this just read what that person stated right? I was sure that Jeff meant any one particular title and that particular article associated with it. If he did not mean this then I do not agree with him but I feel, without even a response from him, that " the title and the article should be consistent" meant exactly what is printed. In that context I still agree. Otr500 (talk) 20:54, 20 May 2011 (UTC)

By reading "a style should be consistent" literally. Thank you for explaining that you didn't mean that, but no more than what JeffConrad said. Septentrionalis PMAnderson 21:13, 20 May 2011 (UTC)

• • But the question for this section is what, if anything, we all agree on; we can work up to a consensus text from there - in Wikipedia's private understanding of consensus, which will still exclude anything to which there is significant active objection. Septentrionalis PMAnderson 17:49, 20 May 2011 (UTC)

From WP:HYPHEN

• • styles do vary among publishers That's the point. I didn't say "consistently dashless," I said "not consistently dashed." Septentrionalis PMAnderson

From WP:HYPHEN_2

• Does anyone deliberately use a minus sign which hasn't identical height, thickness and length as the horizontal bar of the plus sign? (“Deliberately” as opposed to “because they are unable to or can't be arsed to make them the same”.) I'd be somewhat surprised by that. ― A. di M.​_plé​^dréachtaí 03:12, 20 May 2011 (UTC)

  I hate to shock you, but . . . the New York Times Guide to Style and Usage says what it says. And Unix manual pages used a minus sign for an en dash (in partial defense of that practice, original troff did not have an en dash. But the result looked terrible nonetheless.). Several popular books on computer typography that are rigorous on the use of proper quotation marks and dashes seem to think that typewriter single and double quotes (' and ") are used for single and double primes (′ and ″). The moral? As much as I generally believe in following the major style guides, any recommendation still needs a sanity check. Having dealt with these issues when laser printers first became widely available 25 years ago, I’d have thought most people today would be up to speed. Clearly, I was mistaken. I would hope we could do better. JeffConrad (talk) 03:36, 20 May 2011 (UTC)

• Comment; A majority of newspapers are owned by a handful of companies. I do not ever wish to entertain the thought that newspapers are the sole guide to English usage. If this thought is being entertained then just forget everything and follow the New York Times Guide to Style and Usage. Newspapers in my area are horrible at punctuation and possibly saving space may be one intention. I am just stating that I would prefer another guide (primary) than one that can have a broad influence that may not accurately depict the over-all mainstream trend of authors and editors. If an author uses a hyphen or dash and the New York Times has decided against this usage do we agree to do away with them? I am against this as I feel there is a very important need for hyphens and dashes. Otr500 (talk) 13:05, 20 May 2011 (UTC)

  As far as I am concerned, we would permit any reasonable pointing to be used consistently in any article; although systems which are exceedingly rare would probably be changed by consensus. Septentrionalis PMAnderson 18:07, 20 May 2011 (UTC)

• I had found a better verb than require for this, but I can't remember it. :-( ― A. di M.​_plé​^dréachtaí 15:47, 20 May 2011 (UTC)

  God, I hate it when that happens! At least I'm not the only one...
  — V = IR (Talk • Contribs) 17:57, 20 May 2011 (UTC)

  What about justify? ― A. di M.​_plé​^dréachtaí 11:17, 28 May 2011 (UTC)

Exceptions to MOS

Another area of dispute has been around how to handle exceptions. A few options are below. Again, this is not the time for debating these, but it might be helpful to note which have any support or if there are others that I missed. Feel free to remove this entire section if it doesn't belong here, I'm pretty confused about what we're doing so I might be putting this in the wrong place. Note that some of these are not necessarily mutually exclusive. ErikHaugen (talk | contribs) 18:59, 20 May 2011 (UTC)

• The MOS is just advisory or descriptive; any exceptions can be handled at article talk pages, etc.
  • I don't support this at all, I don't want to have the discussion over and over. ErikHaugen (talk | contribs) 18:59, 20 May 2011 (UTC)
  • Support; only really acceptable solution. This is what WP:Policy says guidelines are; the way to not have the same discussion over and over again is to ask the discussants at the second discussion to look at the first. If two or three discussions make the same decision, then it is time to add it to the guideline; that's how guidelines change.Septentrionalis PMAnderson 19:07, 20 May 2011 (UTC)

• Tony's notify-WT:MOS-before-debate compromise: "Where consensus based on a strong majority of reliable sources is arrived at on an article's talk page, an exception to points <regarding use of dashes to mean versus or in other kinds of compounds> may be made for the article and any others that are closely related. The debate should be notified at the relevant talk pages and at WT:MOS"
  • I don't like it, but I'm hoping we can compromise on something like this. ErikHaugen (talk | contribs) 18:59, 20 May 2011 (UTC)
  • Oppose. Only marginally acceptable in Tony's proposal as a compromise; most of the opposition to that proposal is opposition to this idea. Septentrionalis PMAnderson 19:07, 20 May 2011 (UTC)

• Wikiprojects or MOS sub-pages can specify their own exceptions, eg. WP:MOSCHEM might weigh in about "carbon-/–carbon bond" or WP:MOSMED about "basal[ ]/-cell carcinoma", which should be derived relevant style guides and usage in reliable sources.
  • Support. ErikHaugen (talk | contribs) 18:59, 20 May 2011 (UTC)
  • Of course they can. But that does not replace the normal flow of discussion. What a WikiProject can do, any group of editors can do; that's all a WikiProject is. Septentrionalis PMAnderson 19:07, 20 May 2011 (UTC)
  • Example here: "silver(I) fluoride not silver (I) fluoride. Note that this is an exception to the usual English style for parentheses." Art LaPella (talk) 19:12, 20 May 2011 (UTC)

Moving on

1. Does the list of dash disagreements include everything except ranges?
2. Should the authority of the Manual of Style be on the list of disagreements? Obviously we disagree on that issue, so perhaps Casliber should decide the scope of the list he wants.
3. Do we need anything procedural other than a disagreement list, to satisfy the request for "determining the structure of a discussion"? Art LaPella (talk) 17:22, 21 May 2011 (UTC)

   A few thoughts:

   1. It seems to me that we also agree on items 4 and 6 in the current WP:ENDASH (though I’d like to be sure that I understand what’s meant by item 4—the current wording is pretty vague).
   2. If we do agree on items 4 and 6, it seems to me that we implicitly recognize (and accept) differences between common (though hardly universal) practice in the US vs. that elsewhere, and this could have some bearing on some of the items on which we seem to disagree.
   3. We might also find agreement on parts of other items (e.g., Michelson–Morely experiment) if we consider each use separately; I think we should make this separation (and any others that are needed), as Casliber has urged, lest we be left with “partial disagreements” that prove difficult to resolve without voluminous discussion.
   4. I’m not quite sure what you mean by the second item—is this to say that we should decide whether the MOS is prescriptive or advisory for these items?
   5. The question of “consensus” vs. unanimity may also need to be addressed; though WP:CONSENSUS seems impossible if there is any significant disagreement, it’s not clear that it requires unanimity. Moreover, it suggests that consensus is not strictly a numbers game. So far, most of us have simply expressed personal preferences, essentially, WP:ILIKEIT. I realize that a “quality of the arguments” approach could open an unmanageable can of worms; I honestly can’t see anything other than a tabulation of “major” style guides on each point would qualify, which could be a fairly significant undertaking. Hopefully, we could agree on what the guides say, but agreeing on which guides we accept as authoritative could be more difficult.
   6. Whatever we do, I think it’s imperative that we address consistency between the title and article. It seems to me that we agree on this, but if so we should clearly so state, and possibly propose changing the MOS to reflect this.

   JeffConrad (talk) 21:06, 21 May 2011 (UTC)

• Do we agree on items 4 and 6? Well, anybody else?
• OK, if we're separating out "Michelson–Morley experiment", then let's get on with it.
• Yes, "prescriptive or advisory" means the same to me as "authority" (or lack of authority). "on these issues" is a restriction that hadn't occurred to me.
• I thought "consensus" vs. "unanimity" was an easy question. By definition, we need unanimity to say something isn't a disagreement; the only alternative would be to say the other side's disagreement is so wrong that it doesn't count as a disagreement. And we need consensus, not unanimity, when it comes time to settle those disagreements. If the other side is so wrong that their disagreement doesn't count as disagreement, then what better way to get them blocked than to proceed to the debate phase, where they will be obligated to deny the undeniable with arbitrators watching?
• Title matching article? Does that mean choose the title to match the article, and therefore match the MoS, despite what WP:TITLE says? Or does it mean choose the title to match reliable sources, then make the article match that title, despite what the MoS says? Art LaPella (talk) 02:19, 22 May 2011 (UTC)

• It seems to me we may agree agree on 6, but I’d still like to know what 4 is; quite honestly, I think it may be several things, and if that’s the case, it may not be just an instance of 6.
• I’d separate everything that might be a separate issue so that we (hopefully) need only re-vote once.
• We certainly need unanimity to say that there is no disagreement, but we ultimately may need to determine whether there is substantial agreement (or lack thereof) on certain uses.
• By “consistency between the title and article”, I meant simply that do we agree that if the article text uses a dash rather than a hyphen (or vice versa), the title should as well?

JeffConrad (talk) 05:22, 22 May 2011 (UTC)

• • OK with me, except does "substantial agreement" mean everyone except, um, the most active editor in this arbitration? I really don't think that's what the arbitrators' motion means. And I assume your interpretation of "consistency" is a proposed item of disagreement, not agreement. Art LaPella (talk) 05:35, 22 May 2011 (UTC)

Separation of items in a list, etc.

Since no one has offered a better description (or an example), I shall attempt the latter myself. Does “en dashes are used between track titles and durations, and between musicians and their instruments” refer to usage such as that in Kind of Blue? If so, I’d say it is indeed just an alternative to an unspaced em dash (which is used on many music albums that I have). But an en dash is also sometimes used to mark items in a vertical list, especially as a second level under a bulleted list, for example,

• En dashes are used
  • – To stand for to or through in ranges (pp. 211–19, 64–75%, the 1939–45 war). Ranges expressed using prepositions (from 450 to 500 people or between 450 and 500 people) should not use dashes (not from 450–500 people or between 450–500 people). Number ranges must be spelled out if they involve a negative value or might be misconstrued as a subtraction (−10 to 10, not −10–10).
  • – To stand for and between independent elements (diode–transistor logic, Michelson–Morley experiment). An en dash is not used for a hyphenated personal name (Lennard-Jones potential, named after John Lennard-Jones), nor a hyphenated place name (Guinea-Bissau), nor with an element that lacks lexical independence (the prefix Sino- in Sino-Japanese trade).
  • – In compounds whose elements themselves contain hyphens or spaces (the anti-conscription–pro-conscription debate) and when prefixing an element containing a space (pre–World War II technologies, ex–prime minister) – but usually not when prefixing an element containing a hyphen (non-government-owned corporations, semi-labor-intensive industries). However, recasting the phrase (the conscription debate, technologies prior to World War II) may be better style than compounding.

Such a use would obviously not be just a stylistic alternative to an unspaced em dash. A list such as I’ve shown is a bit messy to do even with full access to CSS (which does not include dash as a list-style-type), and is really tedious in Wikipedia, so I’m inclined to assume this was not the intent of “To separate items in a list”. But it conceivably could be read that way by someone familiar with “dash lists”. If indeed usage such as that in Kind of Blue is all that is meant, that or a similar link should make this clear. JeffConrad (talk) 06:25, 22 May 2011 (UTC)

On further thought, using an en dash to set off track times may be more than an alternative to an unspaced em dash (it’s not an application that would occur to me, anyway). So perhaps we should keep it separate just to be safe. As for linking of performers and instruments, I’ve seen this done with commas, unspaced em dashes, and sometimes en dashes; an en dash as an alternative to a comma and space is unusual, so perhaps this also deserves mention. In any event, unless the intent actually was to include a vertical list such as I’ve shown above, the wording should be more clear—in particular, “to separate items in a list” should be clarified, because with the normal definition of a list item, e.g.,

• Item one
• Item two
• Item three

the intended usage does no such thing—rather, it sorta kinda separates stuff in list items. Additionally, we should decide whether linking of performers and instruments requires (bad word, I suppose . . .) an en dash, or whether some of the other more common approaches are acceptable alternatives.

Incidentally, I think it’s unfortunate that a dash list (or at least a mark other than a solid bullet) isn’t available, because a bulleted list nested within another bulleted list can be confusing. But that’s another topic for another time. JeffConrad (talk) 14:20, 22 May 2011 (UTC)

I have interpreted "To separate items in a list ..." to mean that em dashes used to separate items in a list should be changed to en dashes. I can see how someone might interpret "separate items in a list" to mean that the dash comes first, but I interpreted "separate items in a list" as in Kind of Blue#Track listing (but not Kind of Blue#Release history, where the en dashes should be spaced). The em dash section says em dashes are only "parenthetical" and "as a sharp break", not lists. That's my interpretation of the rule as presently worded; it isn't an opinion about what the rule should be. Art LaPella (talk) 16:38, 22 May 2011 (UTC)

And your interpretation is not that different from mine, but the current wording is nonetheless at odds with the normal sense of “list item”, and so vague that I could not understand it without finding an example (recall my original suggestion was simply that an example or a link to one) be provided. The separation of tracks and durations is also reasonably accomplished with a colon (e.g., “So What”: 9:25), and the linking of musicians to instruments as in Kind of Blue#Musicians is most often done, at least on the covers/cases of the 1500 albums I have, with an em dash or a comma (the latter admittedly could be ambiguous in a context such as Miles Davis, trumpet, band leader; a colon would avoid this, though it seems a less common construction.

One could, of course, see what was done with the original. In the case of Kind of Blue, the listing was Miles Davis, trumpet and leader. JeffConrad (talk) 20:41, 23 May 2011 (UTC)

But my point isn’t about what the rule should be, but rather that whatever it is should be clearly indicated, and if alternatives are proscribed, that also should be made clear. Here′s a usage that I can’t find described in any style guide (talk about “eccentric” . . .), so the reader cannot look elsewhere for guidance. A law that requires a person of ordinary intelligence to guess at its meaning is void for vagueness, and I think the same concept is operable here. Again, my main suggestion was a better explanation or a an example (or a link to one); the latter would probably address the issue for now.

It does seem that we’re both saying that this really isn’t a simple case of a stylistic alternative to an em dash, and if that’s indeed the case, items 4 and 6 should not be merged. JeffConrad (talk) 05:29, 23 May 2011 (UTC)

I'm not sure it still matters what we merge, because everyone else seems to have forgotten us. You suggested separate opinions on each example in the guideline (presumably not mine; I never heard of dash rules before Wikipedia; maybe dashes are a British thing.) So would it be helpful if I listed all those examples (excluding ranges, except as spacing examples)?

male–female ratio, 4–3 win, Lincoln–Douglas debate, France–Germany border, diode–transistor logic, Michelson–Morley experiment, Lennard-Jones potential, Guinea-Bissau, Sino-Japanese trade, the anti-conscription–pro-conscription debate, pre–World War II technologies, ex–prime minister, non-government-owned corporations, semi-labor-intensive industries, the New York – Sydney flight, the New Zealand – South Africa grand final, June 3, 1888 – August 18, 1940, June–August 1940, Seifert–van Kampen theorem, Eye–hand span, --

The guideline has no examples for lists, em dash alternatives, nbsp before an en dash, or any kind of em dash. Art LaPella (talk) 17:29, 23 May 2011 (UTC)

I'm not sure we need to break out every example, but rather only each different type of usage, so that we avoid instances of “partial disagreement”. Whether it’s worth the effort depends on what it might accomplish: it seems to me that we might find agreement with Michelson–Morely experiment, and maybe 4–3 win (I also suggested 5–4 decision because it’s a common occurrence, but I think it may be the same type of usage), but I′m not sure we’d find much agreement on other items. It also depends on the ultimate implications of anything but unanimity—what do we do where we don’t have unanimity? Does it matter whether we distinguish between “significant” disagreement and any disagreement?

Because I’ve been the one harping on it, it would seem reasonable for me to follow through by providing the breakout (and I already have a pretty good start on the main Talk; I could simply remove the preconceived results); others could then tweak it as they see fit. Hopefully, we could at least agree on which uses are distinct. But if no one cares, I’m not sure it’s worth doing.

Welcome to the wonderful world of style guide drafting. I’ve been involved in several such efforts over the years, and quickly recognized that it’s more of an undertaking than anyone ever imagines—so much so that I concluded that the only viable approach was to defer to extant, widely used guides and deal only with exceptions or ambiguities in their recommendations. Of course, one must then decide on the “authoritative” guides. In my organization, it was simple—we used what I dictated (For Americans wary of British conspiracies, we relied primarily on CMOS, supplemented by various ANSI and IEEE standards for technical material. Full disclosure: our technical writer was a Chicago alum). In standards organizations in which I was one vote, the process was a more democratic—but CMOS often prevailed because no one had any alternative recommendations. Chicago got there first with the thickest book {nowrap|. . .}}

I’ve used en dashes for 25 years, and thought I understood them fairly well. But apparently my interpretation of what Chicago recommended exceeded what they actually intended; I guess this argues for more examples rather than fewer. And I still don’t really understand item 4; I assume it doesn’t really refer to a “dash” list, but just what does it cover? Only vertical lists? only lists of items that pertain to sound recordings? Does it proscribe the usage on the original jacket of Kind of Blue?

Perhaps the guideline needs examples for lists, em dash alternatives, and at least a few showing em dash. It might be tough to illustrate the effect of   before an en dash without including a forced line break, but the purpose should at least be made clear. JeffConrad (talk) 21:20, 23 May 2011 (UTC)

Additional breakdown of uses

It’s easier to just provide a more-detailed usage breakout than to debate whether it’s worthwhile to do so. If no one responds here, I’ll assume we agree to disagree on everything but items 1 and 6.

I’ve tried to describe a few distinct uses, and have added a few examples to ensure that we all have essentially the same understanding. Feel free to expand the breakout or revise descriptions. I fully recognize the additional complexity, but offhand, I don’t see a good alternative for dealing with the many “partial disagreements”.

As I indicated above, I still don’t really understand item 4:

1. Does it apply to inline lists or only to vertical lists?
2. Does it apply to other than articles on sound recordings? If so, are there some examples?
3. Does the current wording proscribe other ways of separating musicians and instruments (Miles Davis, trumpet and leader)?

JeffConrad (talk) 23:08, 23 May 2011 (UTC) Jeff Conrad

For (3), I'd say commas are fine. Or colons, depending. However, we wouldn't want to use hyphens, which look decidedly unprofessional. I think this was intended as an typographic alternative to hyphens, not a description of whether you should use hyphens in the first place. Or, as others have pointed out, it's really just an alternative to em dashes, and follows em-dashing rules. Not so sure about the others, but (1) I think vertical is the model that they had in mind (remember, commas are acceptable too, and I'd use commas in-line) and (2) sure, operas, plays, etc.: anywhere you give credit. — kwami (talk) 03:55, 24 May 2011 (UTC)

En dashes (–, –) have several distinct roles.

1.

To stand for to or through in ranges (pp. 211–19, 64–75%, the 1939–45 war). Ranges expressed using prepositions (from 450 to 500 people or between 450 and 500 people) should not use dashes (not from 450–500 people or between 450–500 people). Number ranges must be spelled out if they involve a negative value or might be misconstrued as a subtraction (−10 to 10, not −10–10).

• Agree?

2.

2a. To stand for to in a score or vote (4–3 win, 5–4 decision).

2b. To otherwise stand for to (male–female ratio).

2c. To stand for versus (Lincoln–Douglas debate).

2d. To stand for between (male–female relations, French–German border).

3.

3a. To stand for and between independent elements that are proper names (Michelson–Morley experiment).

3b. An en dash is not used for a hyphenated personal name (Lennard-Jones potential, named after John Lennard-Jones), nor a hyphenated place name (Guinea-Bissau), nor with an element that lacks lexical independence (the prefix Sino- in Sino-Japanese trade).

3c. To stand for and between other independent elements (diode–transistor logic).

4.

To separate items in a list—for example, in articles about music albums, en dashes are used

a. Between track titles and durations (Kind of Blue#Track listings).

b. Between musicians and their instruments (Kind of Blue#Musicians).

c. [Other uses and examples?]

In this role, en dashes are always spaced.

5.

5a. In compounds whose elements themselves contain hyphens (the anti-conscription–pro-conscription debate).

5b. When prefixing an element containing a space (pre–World War II technologies, ex–prime minister).

5c. Recasting the phrase (the conscription debate, technologies prior to World War II) may be better style than compounding.

6.

As a stylistic alternative to em dashes (see below).

• Agree?

Spacing

a. Disjunctive en dashes are unspaced, except

b. when there is a space within either one or both of the items (the New York – Sydney flight; the New Zealand – South Africa grand final; June 3, 1888 – August 18, 1940.

c. Exceptions are occasionally made where the item involves a spaced surname (Seifert–van Kampen theorem).

The space before an en dash should preferably be a non-breaking space ( ).

JeffConrad's list ends here

There is more agreement than 1 and 6 alone; for example, we agree on 3b and 5c. We also agree on Spacing, b; but we agree that it's wrong.

We also agree on a restricted form of 3a:

3a'. En dashes can be used to stand for and between independent elements that are proper names (Michelson–Morley experiment) - likewise (2). Septentrionalis PMAnderson 23:48, 23 May 2011 (UTC)

"It’s easier to just provide a more-detailed usage breakout than to debate whether it’s worthwhile to do so" answers my main point. Thank you. Art LaPella (talk) 00:24, 24 May 2011 (UTC)

This seems to suggest that we must further determine whether we agree that a use is required, permitted, or disallowed. I’m not sure we agree that 3a should be optional, and even if we do, what about 3c and all of 5? Are you suggesting we agree that those uses should be disallowed? Though those who have commented here seem to agree that spacing (b) should at the very least be deprecated, we know full well that there are others who strongly disagree, and I’m not sure simply saying “Tough; you should have voted” is a great strategy (and I say this as one who thinks the spacing is off the wall). JeffConrad (talk) 06:43, 25 May 2011 (UTC)

Those who think that Practice P should be required should agree that it is true that P should be permitted. If there is no consensus that P should be required, but a substantial body think so, our options are to permit P or to be silent. Septentrionalis PMAnderson 17:08, 26 May 2011 (UTC)

I agree that there is no real difference between “agreement against” and “disagreement”, but still think we need to distinguish among require, permit, or disallow. That “substantial” agreement short of consensus for requiring a use mandates that we permit (but not require) it is one possible interpretation. But it’s not necessarily the only one—several have indicated opposition to too many options. I should say here that I don’t necessarily disagree with with Pmanderson’s position, but simply acknowledge that it’s not the only one. And we again go back to what constitutes “consensus”. Does it require unanimity? And if not, how far short can it be? JeffConrad (talk) 23:27, 26 May 2011 (UTC)

I have often distinguished unanimity for the purpose of determining the questions in this phase, from the consensus it will take to answer the questions in a later phase. You brought up "substantial agreement". I don't think the difference is important, because of how I (perhaps wrongly) read between the lines of "substantial agreement" and the lines of the arbitrators' motion. To speak more plainly, Septentrionalis's opinions are an underlying cause of this arbitration. If we don't even take his opinions to the debate phase, disregarding them on the grounds that they aren't a substantial disagreement, we would frustrate the main goal of the arbitration. A similar comment applies to Tony1 and spacing. But perhaps you didn't mean that. Art LaPella (talk) 23:54, 26 May 2011 (UTC)

I think this stage is to determine the questions, and the fact that spacing (b) has enough support to remain in the Manual is enough to show that there is significant disagreement on that issue. Art LaPella (talk) 20:43, 25 May 2011 (UTC)

Well, it shows that there is one editor who reverts for it; in this case, I think that's all there is. Agreement against and disagreement have much the same practical effect: MOS should cease to require the spaces. It may or may not permit them; some people will use them anyway. Septentrionalis PMAnderson 17:08, 26 May 2011 (UTC)

Actually, in Wikipedia talk:Manual of Style/Archive 112#Spaces in endash there were at least half a dozen people insisting on the spaces. (Also, it is quite possible that someone could want to e.g. recommend spaces in usage 1, allow both forms in usage 2, and recommend no spaces in usage 3.) ― A. di M.​_plé​^dréachtaí 20:08, 26 May 2011 (UTC)

I think "determining the structure of a discussion on En dashes" means discussing the spaces if we're discussing the rest of WP:DASH (the actual language of the motion is restricted to the title issue.) Even if the only obstacle were one editor who reverts for it, then I would think that would be part of the discussion. If that editor, or any other editor, is obstructing the consensus, then let's start the discussion so the arbitrators can see that behavior here. Are we discussing WP:DASH except "1 and 6 ... 3b and 5c"? And at least the title issue? Art LaPella (talk) 20:18, 26 May 2011 (UTC)

Ostensibly, we probably should limit discussion to uses of the en dash that might conceivably appear in titles. But if we agree that usage should be consistent between the title and the article text (and we seem to), we almost need to address the larger issue lest we end up with a schizoid overall policy on en dashes—I don’t think unintended consequences were the intent of the motion. And the greater issue involves spaces, especially since they could appear in article titles.

OK. Art LaPella (talk) 23:35, 26 May 2011 (UTC)

Concerning titles please comment on these quotes;

• "The phrase "African American" (with no hyphen) is a noun. The phrase "African-American" (with a hyphen) is an adjective. In the past, there has been some confusion in terms of naming Wikipedia articles, but I think they've been straightened out for the most part."
• "Nobody should be confused by "African-American" (with a hyphen) into thinking it has anything to do with relations between Africa and America. That would be signified by "African–American" (with an en-dash)." Otr500 (talk) 13:54, 29 May 2011 (UTC)

Neither is discussing the English language as it actually exists anywhere. Both are efforts to create a dogmatic Newspeak, and if both are genuine quotes from our talk pages (neither shows up on searching), the editor responsible should be ignored until he goes to play on the Newspeak Wikipedia, with its much simpler Manual of Style: "Hyphens are ungood." Septentrionalis PMAnderson 17:44, 29 May 2011 (UTC)

For spaced dashes, here's an example from our shibboleth article. A section heading. Better:

Castilian Spanish-Latin-American Spanish and Portuguese

or

Castilian Spanish–Latin-American Spanish and Portuguese

or

Castilian Spanish – Latin-American Spanish and Portuguese,

in a context were all headings are X-Y / X–Y.

BTW, what we had here, and what I've seen on many articles, were spaced hyphens: X - Y. — kwami (talk) 02:03, 30 May 2011 (UTC)

May 30

May 30 is tomorrow. The consensus seems to be to ignore the arbitrators' distinction between determining questions and answering them. OK, I'm assuming the new plan is to determine consensus based on the questions above. Art LaPella (talk) 20:52, 29 May 2011 (UTC)

I think we still need to state what we think the questions are, even if we’ve answered them. If we have agreed on the questions, stating them should be simple and without controversy. But if there is disagreement about any of the stated questions, the answers (which in many cases, are that we disagree) are suspect.

At this point the main questions seem to be whether we agree or disagree on en-dash uses in the current MOS. But I think we also need to include

1. How prescriptive the MOS should be.
2. Whether the MOS covers titles, and if it does, to what extent.
3. Whether we think usage in the titles should be consistent with usage in the text.
4. If we agree that, when a usage finds significant support short of unanimity, we agree that the usage should be permitted but not required (assuming, of course, that the MOS is at least partially prescriptive).
5. What we do with the MOS if we find we are short of unanimity (and perhaps even consensus) on some uses. Would this make the current MOS a non-consensus policy, and consequently render it null and void? Or would we conclude that, lacking consensus for change, the status quo prevails?

JeffConrad (talk) 22:08, 29 May 2011 (UTC)

Jeff: all good questions, but rather more major in scope than I think anyone thought of for this. Here are brief responses.

(1) All style guides are both prescriptive and descriptive. To be descriptive alone means to recommend nothing where usage varies: all style guides strive for best practice. To guide is to mark some usages as preferred, and some as deprecated.

(2,3) There's very little overlap between WP's Manual of Style (all 30 or so MoS pages) and WP:TITLE, which says nothing about punctuation, for example. The talk pages of WP:TITLE and WP:MOS – perhaps also some other MoS pages – need a short reminder at the top about coordination in matters of article titles and article text.

(4,5) No professional publisher (in print or on the web) allows random, disparate style. Unanimous support for every clause is rarely possible on editorial boards; and no one here agrees with every point at WP:TITLE, or at the MoS pages. The usual consensus through robust discussion has served us well, although we might need small improvements in procedure.

I want to make it clear that harmony among editors should be the first priority. I've no intention of getting into any steamed-up debates about en dashes; nor do I intend to go around systematically changing punctuation in article titles. Where reliable sources truly are a relevant consideration, that should be noted. And if there's strong support among a community of editors for an exception for the whole "Mexican~American War" group, sure, I'll make no objection. An editorial note at the top of each affected article would be helpful for newcomers. Tony (talk) 08:37, 30 May 2011 (UTC)

I have a problem with the "An editorial note at the top of each affected article would be helpful for newcomers." approach, because I see that used as a mechanism to "lock in" one particular style or other regardless of any outside opinions. I find that sort of "solution" to be worse than the perceived problem, especially since there really isn't a problem. There's no need to attempt to prevent styles from changing over time, after all. Besides, there's no requirement to check a pages talk page prior to editing the page. Let's not build artificial barriers, please.
— V = IR (Talk • Contribs) 12:09, 30 May 2011 (UTC)

I agree that harmony among editors should be paramount, so that the energy devoted to arguing (sometimes vehemently) about en dashes can be applied to more productive ends. My reason for suggesting that we include some of these points in a summary of this discussion is to make rehashes of the same issues in the near future less likely.

(1) If we agree that the MOS (in this context, which is all we were asked to address) is in some cases prescriptive, advisory in others, and perhaps even simply descriptive in others, then our summary of this discussion should so indicate. If it is concluded that some changes to the MOS are indicated, and we can apply these terms to specific uses under WP:ENDASH, so much the better.

(2),(3) If we agree that usage in titles should be consistent with that in article text (and I think we do), we should simply so state. Recall the hyphenation of ass-u-me . . .

(4),(5) The operative word is random. Most publishers are pretty fussy about consistency within a single work, but many, including Chicago, are somewhat flexible about accommodating individual authors if those authors are self consistent. Complete inter-article consistency would be nearly impossible even with consensus (which we probably could never achieve); if we conclude that certain uses are required, encouraged, permitted, deprecated, or proscribed, we should simply so state (even if most are simply permitted with consistent usage). By doing so, we hopefully minimize the number of lengthy discussions such as this in the future.

Succinctly summarized: indicate what we have agreed to agree on, so we don′t need to discuss it again next week. JeffConrad (talk) 23:47, 30 May 2011 (UTC)

[Please comment after my post, not within it.–Noetica]

Art points out that questions need to be finalised. Jeff has posted useful ideas about the wider context in which MOS guidelines operate. But this page has returned, over its eleven days, to the MOS guidelines themselves. Four big questions emerge:

1. How much support should MOS have from style guides?
2. How much support does MOS have from style guides?
3. Which guides (or what general sorts) are relevant?
4. What are the best protocols for establishing guidelines?

From the wealth of discussion above, answers emerge also:

1. MOS, it seems to be agreed, must respect major style guides, and must show that it does so.
2. MOS already accords well with major style guides (going by the evidence for the en dash). If anything, it agrees with them too much; it's more eclectic than any other guide I've seen (except perhaps Cambridge Guide to English Usage). Nothing else, for example, has those limited-use spaced en dashes (1 July 2009 – 3 May 2011) and en dashes like the ones CMOS wheeled out a century ago (Ex–Secretary of State). MOS already caters for diversity, so there's something in it for everyone to dislike. Efforts to please one contingent must displease another group. Robust style guides live with that, and so do their users – just as we live with the awkward inevitability of π being irrational. It's how things are.
3. There is a core of respected style guides to take seriously. I see little support for newspaper stylebooks (APA's, the Associated Press Stylebook, for example). Their task is different. And there are no serious guides (beyond WP:MOS itself) for enormous collaborative online projects, so MOS must turn to the big names used for print: CMOS, New Hart's Rules (with its adjuncts), Butcher's, the better M-W products, the Gregg Reference Manual (huge and meticulously detailed; rarely mentioned here), and a few others. Then there are scientific and journal-publishing guides: the superb ACS Style Guide (from the American Chemical Society), and Scientific Style and Format (the CBE Manual). Preferences will differ; but it is agreed that these are weighty, while Truss's travesty is not.
4. Guidelines are to be negotiated by orderly discussion. Surely! No one seriously calls for a mere show of hands; no one expects (or gets) guidelines by fiat; and no one seriously expects unanimity. If those developing MOS have fallen short of high standards, let's do better. Let's be more vigilant to stem lapses when they arise.

Is any of that outlandish? Doesn't it arise naturally, from the tide of discussion on this page?

I agree with Tony's points above, though I would stress different notes in the same tune:

• (1) MOS does have to "prescribe"; but it's a separate issue, beyond MOS itself, to determine how its recommendations will be received.
• (2,3) WP:MOS is the Project's source of guidance for punctuation. No other guideline or policy page deals with it, including WP:TITLE, which is equally accepted as the place for specific policy on titles. Readers will expect title and text to match: in spelling, in form of name – and in punctuation where that is relevant. But there is no "jurisdictional" clash: WP:TITLE correctly stresses reliable sources; nothing in WP:MOS contradicts it. The logic is straightforward: WP:MOS could not make recommendations on punctuation (WP:HYPHEN, etc.), if WP:TITLE magically settled punctuation of titles by comparison of reliable sources (which vary wildly in punctuation), and therefore (because consistency is needed) in article text as well. No, WP:MOS has always advised on hyphens and dashes in titles; and if its guidelines are carefully developed, properly scrutinised, and clearly expressed, there is no genuine problem.
• (4,5) MOS should certainly take note of best practice (formulated in serious style guides), and present clear, usable standards for Wikipedia. If it achieves that, editors will accept it more happily. Otherwise they will not, and there is nothing anyone should, or can, do about it – beyond striving for an even better MOS. In the end, any style guide will cause grumbling; as much as anything else, this shows that editors are indeed guided by it, toward making better articles. And that means a better Wikipedia.

Is there any question to take to the community, now? I struggle to see how it would go, or what the point would be. I see no will for it. Taking Tony's lead, I too can concede the Mexican~American War articles for the foreseeable future, since it means so much to some editors. Let's agree to consistent use of hyphens in them all. ArbCom could move to settle that matter, or editors on this page could endorse a multiple RM that no one here should oppose.

Meanwhile, how about a return to normal development of WP:MOS. It's been protected for three months, and there's a backlog of barnacles to scour away. What's more, there's now a great deal of excellent commentary (some on this page) to use in polishing the hyphen and en dash guidelines. They're in good shape, but let's accept opinions expressed here: the details need to be articulated even better.

It's time to finish this. Let's learn from recent troubles; and let there be a return to core business – with more respect, more focus, more acceptance of diversity, but also more acceptance of strong, well-founded guidelines for the sake of all the articles, and all their readers.

Noetica^Tea? 15:05, 30 May 2011 (UTC)

I agree that absent a very good reason for doing otherwise, we should be in accord with major style guides (and for the most part, I agree that we are); I’d probably include a couple of others (e.g., OSM and Garner’s Modern American Usage) in the list, but that’s more of a detail. The folks who have developed these guides aren’t exactly blockheads, and needlessly reinventing the wheel seems a questionable use of time and effort.

If we decide that, for now, we simply choose to deal with the titles issue that got ArbCom involved, we should so state, perhaps concluding

1. There is some flexibility in choosing between en dashes and hyphens, especially when quality sources disagree, and the decision on which to use is to be worked out among an article’s authors in an orderly discussion.
2. That use of dashes and hyphens in a title be consistent with usage in article text (perhaps with title redirects regardless of what is chosen for a title).

Again, whatever we’ve concluded here, we should summarize as succinctly as possible. No person in his or her right mind is going to sort through all that’s been said. JeffConrad (talk) 00:11, 31 May 2011 (UTC)

Anybody who claims that MOS accords with most style guides on dashes is presenting deliberate falsehoods. Our section on further reading chooses eight style guides. One of those supports the present wording; one is inaccessible to me; the other six do not. Similarly, a much longer list of style guides was compiled from Amazon, in order of sales rank. Of those, much less than half support the present text - and all of those but one are from the lower half of the list; including all those Noetica now cites. I therefore request that Noetica retract that claim, and all that depends on it; otherwise, I intend to cite this post as part of the evidence that Noetica is intellectually dishonest, and should be topic-banned.Septentrionalis PMAnderson 21:02, 1 June 2011 (UTC)

Hmm, sounds like we missed a point of disagreement. It also sounds like at least one editor needs to be told to stop saying things like that. Art LaPella (talk) 21:54, 1 June 2011 (UTC)

I am very tired of a small group of editors who will say and do anything to get their way. Septentrionalis PMAnderson 22:01, 1 June 2011 (UTC)

Yes, I know that is your position. Septentrionalis may feel free to interpret "at least one editor" as Noetica. And vice versa. Art LaPella (talk) 22:16, 1 June 2011 (UTC)

Then I wish your comment good fortune. And presumably Noetica does too. Septentrionalis PMAnderson 22:21, 1 June 2011 (UTC)

I wish we could avoid personal attacks and contentious comments such as “deliberate falsehoods”. Be assured that we have far more than one editor with a position not terribly different from Noetica’s. I remain baffled by some of Pmanderson’s statements about which style guides support what; I repeat here last week’s list of the top sellers on Amazon’s US web site:

1. Publication Manual of the American Psychological Association, Sixth Edition: #19

2. The Chicago Manual of Style, 16th Edition: #449

3. The Elements of Style (4th Edition): #526

4. A Manual for Writers of Research Papers, Theses, and Dissertations, Seventh Edition: Chicago Style for Students and Researchers: #696

5. Garner’s Modern American Usage, 3rd ed.: #2,074

6. Merriam-Webster’s Guide to Punctuation and Style: #12,396

7. The Chicago Manual of Style, 15th Edition: #14,386

8. MLA Style Manual and Guide to Scholarly Publishing, 3rd Edition: #20,161

9. The New York Times Manual of Style and Usage: #27,743

10. Words into Type, Third Edition: #34,474

11. The Associated Press Stylebook: #111,881

12. The Cambridge Grammar of the English Language: #122,589

13. The Chicago Manual of Style, 14th Edition: #151,441

14. Fowler’s Modern English Usage, 3rd rev. ed.: #191,507

15. Fowler’s Modern English Usage: the Classic First Edition, ed. by David Crystal (2010): #194,870

16. Scientific Style And Format: The CSE Manual for Authors, Editors, And Publishers: #238,629

17. The Oxford Style Manual: #297,890

18. New Hart’s Rules: The Handbook of Style for Writers and Editors: #406,432

19. Merriam-Webster’s Manual for Writers and Editors: #796,570

As nearly as I can tell, (5), (6), (16), (17), (18), and (19) are in almost complete agreement with our current MOS (save perhaps the spaced en dashes), and (2), (7), (10) support many of the uses (I don′t have (13), but assume CMOS 14th is similar to the 13th, 15th, and 16th). Guides (3) and (14) are silent on the issue, and guides (1) and (8) are largely silent, failing to mention even uses they employ; it has been suggested that silence is disagreement, but I take that as simply one editor’s opinion. Guides (9) and (11) specifically deprecate the en dash (the NYT guide deprecates use except as a minus . . . need I say more?). I don’t have access to (4), (12), and (15), so I can’t comment.

I agree that that it would be incorrect to claim that most major style guides support almost everything in our current MOS. But to imply that the current MOS finds little support in the guides is just not supported by the facts. JeffConrad (talk) 22:51, 1 June 2011 (UTC)

Yes, guides vary. While I was away, 14 cheap used guides (less than $7 with shipping) that I ordered showed up at my house. Seven of them don't mention en dash at all, and seven of them overlap our guidelines to varying extents. Some specifically note that the en dash is problematic for not having a represenation on the typewriter, or that authors don't use it in manuscripts but that editors mark en dashes on the way to the typographer; the guides for editors and typographers mention it more than guides for authors do, in consequence. If you look at the Cambridge grammar (largely visible on Amazon, too rich to buy), it's clearly the most in-depth, sophisticated, and descriptive as opposed to prescriptive. It describes the extent to which usage varies (a few of the others do as well, noting that in some styles hyphens are commonly substituted for en dash), and it discusses British versus American usage (which is just in the use of spaces around en dash with spaced compounds, which is British, not American). It refers to the punctuation role of the en dash as "long hyphen", being a more descriptive term for its use, and explains that the long-hyphen indicator is realized by either an ordinary hyphen or an "en-rule", depending on the style. This supports our use of en dash as a typographic style, not a difference in name or spelling that could possibly be in conflict with WP:COMMONNAME, which was one novel theory against our style guidelines.

It's interesting to see the different authors of guides infer the rules from the usage, some pretty general, and some pretty specific. I've been taught the more general form, so to see it broken up into special cases like our MOS does seems odd, and invites nit-picking; so I agreed that the merge of the different spaced en dash as em dash forms would be OK; I'd also merge some of the cases, and just go with the more general interpretations like the ones I was taught: "used between words of equal weight in a compound adjective", "used when the two elements are equal in importance and may be reversed without altering the meaning", "to link a compound term serving as an adjective when the elements of the compound are parallel in form (that is, the first does not modify the second)", "when the first part of the compound does not modify the meaning of the second part", "can be used to link 2 words or phrases representing items of equal rank and 2-word concepts", "semantic relationship is 'between X and Y' or 'from X to Y' ... [also] more than two components ... also found with adjectives derived from proper names", "to express tension or difference, or to denote a pairing in which the elements carry equal weight (love-hate relationships, contract-tort doctrines)", "when you have an equal-weighted pair serving as an adjective", among others. Yes, many guides also allow, or admit the use of, hyphens, saying things like "(Some publishers accept a hyphen but those who use the en dash feel that a hyphen is a poor substitute and should be reserved for its own distinct uses, as a connector in compound words and as a separator in showing syllabification.)" Our MOS is based on all of these concepts, however expressed.

I was pleased to see the broad agreement. To the extent that we have a few people who take issue with some details, we can entertain proposals for changes; but for such disagreements to be taken elsewhere as a basis for ignoring and undermining the MOS seems to be a concept that the community has little taste for. We spent a few weeks to learn that one or two people disagree with a couple of details of the MOS, in exactly the places that were obvious at the outset. Let's be glad that we don't have a situation where we have to chuck it all and start a large-scale restyling, undoing the work of the last six years that brought WP up to a pretty good and consistent styling. If we keep in mind the permit/require distinction as I discussed above, it should be clear that no editor or reader has been inconvenienced by these provisions of the MOS, nor will they be in the future. If those editors who previously noted that they can't see the difference, or don't care, will refrain from stirring the pot, we can expect this stew to come out good.

My thanks to all who participated, and my apologies for being rather unavailable to help while on the road in Eastern Europe and Asia. I took a photo of an English explanation of artifacts in the Archeological museum in Istanbul, to show how they used the en dash and hyphen there, in conformance with our MOS – unfortunately, I had my camera and computer stolen in London yesterday, so I can't share that with you, and I don't recall the examples I found; but they're not unique to my American training, nor to Britain.

Dicklyon (talk) 03:53, 2 June 2011 (UTC)

From the arbitrators' point of view, it doesn't matter if the rest of Wikipedia cares about style manuals, because that isn't where Septentrionalis just drew his line in the sand with everybody watching him. Art LaPella (talk) 05:35, 2 June 2011 (UTC)

Well, it wouldn't be the first, second, or third time he has accused me of lies and deliberate falsehoods, but I speak in good faith about what I have found, from looking way beyond the list of 8 that he refers to, and can provide copies and quotes from these many manuals to anyone who wants to dig deeper. He makes a good point, though, that our further-reading section could be expanded to better cover the sources of the en dash usages that we have adopted. Dicklyon (talk) 13:22, 2 June 2011 (UTC)

I must conclude that some few editors are so enamoured of the Oxford dash, that they are simply incapable of reporting what they see. Let us take the post that begins this section:

• Is this list of 19 books all of Amazon's style guides? By no means. Amazon has 1672 hits on "writing style guide", most of them what we are discussing as style guides.
• Are they the top 19, or any other random selection? By no means. To suppose that would be to suppose that there were only 15 style guides in the top 194,870 books, and only 4 more in the next 600,000. (Not only absurd, but demonstrably false.) No, they are close to the top 14 style guides (of which two support the Oxford dash) salted with five poorly-selling ones, all of which do.
• Does CMOS (16th edition) support "many of the usages" in the present text of MOS? Disingenuous. It supports the uses on which this conversation agrees; it expressly does not support US–Canadian relations (or any use of an en dash as between); it is much more cautious than the present text of #5, and even so expressly supports non-English-speaking peoples and a two-thirds-full cup with two hyphens. (And thinks the latter should be reworded; as it should.) That leaves passing permission of part of #2, and part of #5. Septentrionalis PMAnderson 17:33, 2 June 2011 (UTC)

  I think most of us know how to conduct a search on Amazon (or whatever); the hits are obviously more comprehensive, but require considerable culling to arrive at a useful list. I don’t think anyone suggested or imagined that the list above comprises all extant style guides; I simply included many that I have used for years, and other for which inclusion seemed indicated. But the list was originally posted on 13 May, so there has been plenty of time to address any significant omissions (in retrospect, I clearly should have included the MLA Handbook). If there are suggested additions. please feel free to make them. If someone starts citing the Hillbilly Gazette Gide to Gud Style (or its UK equivalent), I think it’s quite reasonable to cry foul. But I don’t think we’ve done anything of the sort

  I think we readily acknowledge that CMOS 16 specifically does not endorse the en dash in the sense of between. I don’t think we disagree on your specific summary except perhaps on how to label it. JeffConrad (talk) 00:04, 3 June 2011 (UTC)

  Very well, you deny you have cherry-picked obscure style guides from OUP (and perhaps some other publishers whose house tradition is to publish style guides they do not themselves follow). The evidence does not support you. Septentrionalis PMAnderson 16:48, 3 June 2011 (UTC)

  Were it not a violation of talk page guidelines, I would respond with the final example of en-dash usage in the OSM. Though I don’t expect you to agree with me, I take issue with being called a liar. What’s puzzling is that the accusation doesn’t even make sense. Unless I’ve missed something, only (5), (14), (15), (17), and (18) are from OUP. I think we agree that (17) and (18) are not common in the US, but (17) is quite common elsewhere. To suggest that MAU and MEU are “obscure” is simply absurd. And to imply that Garner is somehow an OUP stooge is as insulting as it is absurd. So I guess I fail to follow your reasoning.

  For what it’s worth, I “cherry picked” most of the guides listed by including everything on my shelf (of the 19, I have 12), and originally cherry picked most of those by getting everything on the shelf of local bookstores (in the days when there were local bookstores). Except for New Hart’s Rules and USM, my local library has almost the same set of guides as I do, so we apparently share a penchant for obscurity. I added the AP and NYT style guides because of Dank’s comments, and you added the original edition of MEU.

  Again, I don’t think anyone claims that the list above is complete or even the best 19-guide selection. But the suggestion that the list was cherry picked in furtherance of some OUP conspiracy is a red herring that merits no serious consideration. JeffConrad (talk) 02:21, 4 June 2011 (UTC)

  And I object quite strongly to having an unnatural usage supported by irrelevant citations of a loaded list. Modern English Usage is not obscure; indeed, the second edition, by Gowers, outsells much of this list, despite being out of print (it preserves much of Fowler's text, and agrees with him where it does not). Burchfield's redone edition, which saves nothing but the title, is obscure - and deserves to be. Septentrionalis PMAnderson 23:42, 4 June 2011 (UTC)

  Full disclosure here: I have the 2nd ed. of MEU and prefer it (Do we actually agree on something?). I also have the 2nd ed. of MAU, though the differences there are far less. As of a couple of minutes ago, Amazon US showed the 2nd ed. of MEU at #544,407 in Books, so if we take that ranking as meaningful (and I’m not sure it is, especially for an out-of-print book), the 2nd ed. is far more “obscure” than the 3rd. That you think it should be just the opposite does not affect the ranking.

  I don’t know quite what to do with “unnatural usage supported by irrelevant citations”, but let’s try to address the issue of a “loaded list”: what “unloaded” list would you propose in its place? Dank’s approach of taking the top sellers by category is arguably better than basing the choice on one’s own collection; if nothing else, it should make the choices less capricious (at least to the extent that Amazon’s categorizations aren’t capricious). This approach doesn’t directly give the overall rankings, but they could be readily found. One could also look at the bibliography in CMOS, which is quite comprehensive (though I concede it doesn’t include everything in the list above), or at 'OSM (which gives a much shorter list), or some other guide from a major publisher. JeffConrad (talk) 00:26, 5 June 2011 (UTC)

  • No, I prefer the first edition; the second is worth looking at because it preserves much of Fowler himself.
  • One unloaded list would be the top twenty style guides; this would be an arbitrary choice, but at least random. It would, it appears, include two proponents of the Oxford dash, printed by OUP and Merriam-Webster. That neither publisher follows its own traditional style guides is a strong indication that we should not either. Septentrionalis PMAnderson 21:44, 5 June 2011 (UTC)

    But even a list of the “top twenty style guides” is very dependent on how it is obtained. On the US site, almost any relevant search will yield many extraneous results unless it’s constrained to “Reference > Writing”, and even then different results are obtained for style and style guide, and the more-restrictive latter criterion still generates quite a few extraneous results that include multiple editions of the same work, works essentially derivative of other works in the list, and works that are essentially irrelevant. Moreover, the overall (Books) rankings seem to change considerably in the course of a few days. For the UK site (yes, there actually are other English-speaking countries) it’s even more complicated because of the additional subcategories. And the rankings within subcategories don’t always match the overall rankings (e.g., Butcher is a step above CMOS in most cases, yet the CMOS overall ranking is far ahead of Butcher).

    So I’m not sure it’s that simple to get an unbiased, “authoritative” list. Though looking at Amazon rankings seemed an improvement over arbitrary claims of “top” style guides, I nonetheless indicated my reservations about the significance. Nothing in the course of further investigation has led me to feel otherwise. JeffConrad (talk) 01:06, 6 June 2011 (UTC)

  I would be perfectly willing to accept a top twenty from Amazon UK; it would only prove that the extensive use of the dash is a minority view in England, as well as the US. Septentrionalis PMAnderson 00:38, 7 June 2011 (UTC)

Re-commenting in split areas

Can everyone who has already commented at Wikipedia talk:Manual of Style/dash drafting on questions that have now been segmented please go and comment in the individual segments? Show the way for the flood of other editors. Apologies for making folks do some segments again. Cheers, Casliber (talk · contribs) 05:26, 4 June 2011 (UTC)

I admit I was annoyed, thinking that the results were already clear and this was unneeded extra rehashing. But as I see the comments developing on the more finely divided points, my feelings changed. You have managed to elicit a lot of detailed constructive ideas for improvements. Thanks. Dicklyon (talk) 03:22, 6 June 2011 (UTC)

Mexican–American War

Does 2.d. cover this? "[Dashes are used] To stand for between (male–female relations, French–German border)." ErikHaugen (talk | contribs) 19:37, 4 June 2011 (UTC)

It would appear to be a separate "rule". New Hart's Rules (the most common style guide I know of to endorse that against usage) makes it a separate sentence. Needless to say, there is no agreement on it. Septentrionalis PMAnderson 23:34, 4 June 2011 (UTC)

Among the usages that I've found described in guides that would seem to encompass Mexican–American War and French–German border (which do both appear that way in multiple books) are:

• to mean the equivalent of "and", "to", or "versus" in two-word concepts where both words are of equal weight
• used when the two elements are equal in importance and may be reversed without altering the meaning
• to link a compound term serving as an adjective when the elements of the compound are parallel in form (that is, the first does not modify the second)
• when the first part of the compound does not modify the meaning of the second part
• semantic relationship is "between X and Y" or "from X to Y" ... [also] more than two components ... also found with adjectives derived from proper names
• to express tension or difference, or to denote a pairing in which the elements carry equal weight ... wherever movement or tension, rather than cooperation or unity, is felt
• use en dashes when you have an equal-weighted pair serving as an adjective
• items of equal weight
• used to indicate linkages such as boundaries, treaties, and oppositions
• often equivalent to "to" or "versus"
• to denote a pairing in which the elements carry equal weight
• between words that retain their separate identity in a compound adjective

Dicklyon (talk) 02:54, 6 June 2011 (UTC)

What's best for Wikipedia

I'm having a tough time staying away from this. :)

Anyway, looking over everything that's been said, it's obvious to me that everyone involved in this is well informed on the issues (some more so then others, I'm sure, but we all seem well enough educated about English to be able to talk about the subject). There is voluminous debate here about the existence and merits behind various books on English grammar, but I don't see much discussion about what is best specifically for Wikipedia. While our medium is hardly unique, there are unique aspects to it. We can certainly agree that Wikipedia isn't a print medium, I'd hope. So, I'm hoping that along side continuing to spin our wheels and provide ammunition for the arbitration committee by going back and forth at each other about the merits (or lack thereof) to various grammatical works, maybe we could start a discussion about the reasons that certain practices are good or bad for Wikipedia.
— V = IR (Talk • Contribs) 01:05, 5 June 2011 (UTC)

Sounds like a can of worms that may prove unclosable once opened . . . The best thing for Wikipedia, of course, would be to put the time we’ve invested here into fixing far more serious shortcomings in articles (e.g., lack of sourcing and sometimes even content). But what could we do?

1. We could make the decision on a criterion such as following style guides (which don’t always agree), following usage (for which it could prove difficult to present a convincing case), consensus (our normal approach, though finding it here is proving difficult), or editors’ preferences by sheer numbers (which is not the way we normally do things). But how would we choose the criterion while avoid the current seemingly endless discussion?
2. We could avoid non-keyboard (essentially, non-ASCII) characters across the board, as some have proposed and avoid the problem altogether (many online version of newspapers do this for various reasons), though this might be difficult to reconcile with thousands of articles that already include non-ASCII characters, and would make the English Wikipedia almost unique. And it would pose serious problems for many technical articles.

I guess this is to say that I don’t have a ready answer.

Though I certainly agree that Wikipedia isn′t a print medium, I would ask, “So what?” And how many different kinds of “non-print media” are there and how do they relevantly differ? I suppose for an online medium, the display medium (e.g., a smart-phone display) might be a consideration. But if we were to seriously examine this, I think we’d find that dashes, quotes, and similar typographical details are a small part of the issue. And I think we should recognize that content can and will be repurposed, quite possibly for another medium, so we would be well advised to avoid writing the program that tracks exactly seventeen salesmen. JeffConrad (talk) 01:38, 5 June 2011 (UTC)

Well, one thing that comes to my mind is that quite a bit of this issue is about typesetting. I mean, regardless of what we decide, or endorse, or whatever, we obviously don't and shouldn't expect people to enter the "correct" dash character from jump. That, and there's no locking articles once they're "done" (which they never actually are). I don't know, all of this is pretty nebulous in my own mind. I'm not one to lean on others much though, so I don't see any need to be enslaved to exactly what various style guides say. We should certainly be informed by them, but... as far as I know, none of them are written with the environment that we deal with here in Wikipedia in mind. I see the above sections moving too (if y'all aren't already there) some sort of a style guide battle royal, counting up the advice that they give and trying to weight them all... I don't see that as being a good way to settle all of this.
— V = IR (Talk • Contribs) 00:38, 6 June 2011 (UTC)

I agree that the current process is not ideal; the MOS has served us pretty well, and seems to enjoy broad support among people who know about it. Most editors remain oblivious, and enter all kinds of random junk that other editors have to clean up; such is life. It baffles me that a big deal is being made of this in the case of a particular minor punctuation issue. On the bright side, it appears that the detailed breakout that Casliber has assembled has attracted some very constructive ideas on how to fine-tune the en dash section, merging, simplifying, and rebalancing some of the points. It clearly won't be possible to make everybody happy at once, but it looks like we can probably find a broad consensus for modest changes that will improve the MOS. Should we start to propose and edit a replacement section, now that it's becoming clear what to do? Or wait six weeks?

As for arguing over books, that is indeed silly. In my estimation, half of all writer's guides don't mention en dash at all, because their audience is people with typewriters, or with non-en-dash styles, or for writers who will leave the typographic style up to their editors. We already do accomodate that approach in WP, by allowing editors to just use hyphens, and other editors to style it toward MOS standards. So there's no sense in trying to "average those in". The only real disagreement seems to be the Chicago versus American/British/Australian usage. But the choice that was made years ago seems to be widely supported among WP editors who know and care about en dashes, and irrelevant to those who can't see the difference or don't care. Certainly the distraction caused by a few who very strongly prefer the Chicago style has been disruptive, but we'll get through it and back to constructive editing, I expect. Dicklyon (talk) 03:17, 6 June 2011 (UTC)

I′m somewhat baffled by “Chicago versus American/British/Australian usage”; unless I′ve really missed something, Chicago style is American usage, if one is to judge by sales and citations in other style guides. It’s certainly where I start, but I don’t think it’s necessarily dispositive on every issue, even in the US. Though it′s probably silly to argue about which guides can fit on the head of a pin, I think it′s equally silly to say they should be paid no heed. In the US, I would have a hard time with a list that did not begin with CMOS and include at least APA, MLA, S&W (though it’s not of much help here), probably GMAU, and perhaps CSE and a few others for specialized applications. In the UK, I would be suspicious of any list that did not include Butcher, OSM (or at least NHR), and yes, CMOS and S&W. We could argue forever about some of the others. I’ve used MWM and its predecessor M-W’s Standard American Style Manual for years because it covers many relevant topics and does so succinctly; whether it should be included in a short list is probably an unanswerable question.

I′m also surprised by the seeming deprecation of CMOS. Though Chicago don’t support every usage in the current MOS, they’re hardly dash averse, as a glance at these entries from their monthly Q&A indicates. They really don’t differ that much from OSM, save perhaps the sense of between (perhaps better termed and). Chicago’s approach differs primarily in that it uses en dashes only when essential; this Q&A indicate that Chicago staff like subtle distinctions such as the en dash provides, they feel (probably correctly) that most readers will nonetheless see a hyphen. The result is that they would use US-British relations but apparently United States–British relations because of the open compound. Each to his own, but I find this somewhat confusing (something is different . . . why?). And again, for those who can’t see the difference, what’s the problem? Of course there is the distinction between Arab–Israeli conflict and Arab-Israeli conflict (the latter possibly a domestic dispute involving an Israeli couple of Arab heritage?). Had Mitchell been charged with determining whether the issue should be described using an en dash or a hyphen, he probably never would have even accepted the assignment.

Though I support most of the en dash usage in our current MOS, I still consider myself very much a proponent of Chicago style. JeffConrad (talk) 05:21, 6 June 2011 (UTC)

Jeff, I have nothing against the Chicago style, either; it's just that the CMOS keeps being cited by a few as the reason to ignore one wide class of usage that's not represented therein, but is widely represented in other American, British, and Australian guides. It doesn't seem appropriate to call this the "Oxford dash", since Oxford's guides are mixed on it, and many other publishers describe it as well (and also because the term Oxford seems to suggest that it is not in guides of American publishers); so I took Chicago to be the distinguishing feature of the argument. Dicklyon (talk) 06:04, 6 June 2011 (UTC)

Dick, I don’t think we significantly disagree. GMAU, the obscure MWM (as well as the less obscure but also less comprehensive MWG), and to an extent, APA are the only guides that support most of our current WP:ENDASH, though many of the others are simply silent on some of the uses. Chicago specifically disagree with one usage, so they’ve apparently been libeled. As for my reaction to every reference to “Oxford dash”, further discussion would violate talk page guidelines. JeffConrad (talk) 06:43, 6 June 2011 (UTC)

I’m still a bit unclear on “the environment that we deal with here in Wikipedia” as it relates to style guides. Do you simply mean that many editors are new to writing and unfamiliar with such guides? If so, I have no disagreement, but wonder where it leads? Is it to suggest that anything goes? As Dick mentioned, we don’t get bent out of shape when readers enter hyphens for dashes, ASCII single and double quotes for primes, and similar—other editors serve as copy editors. Punctuation is easy to fix; unsourced garbage is a far more serious shortcoming that also often requires far more effort to fix. What surprises me are the vehement objections to cleanup that arguably changes what’s acceptable for many people to the proper usage (or typography if you prefer). For those who argue (probably correctly) that most readers cannot tell the difference between an en dash and a hyphen, what then is the problem with using an en dash?

It’s sometimes tough to distinguish typography from usage. Certainly, with access to many more characters, issues of usage arise that did not exist with a typewriter. But this additional repertoire has also led to some problems; uses (such as a hyphen for an en dash or minus sign) that gave an acceptable appearance with a monspaced font look far less acceptable (and sometimes result in ambiguity) in typeset material. So I think it’s more than a matter of being fancy. JeffConrad (talk) 05:21, 6 June 2011 (UTC)

[Please: comments beneath my post, not within it. Noetica^Tea? 07:27, 6 June 2011 (UTC)]

Jeff:

First let me thank you for your detailed and diligent work on these pages. It's time for me to comment on matters you touch on above.

1. Weightiness of style guides

I appreciate this business of listing Amazon sales numbers, but let's stress once more how incomplete that is as evidence of importance in serious publishing practice. We aren't sure how they are derived, for a start. Are they historically accurate, or only recently gathered with care? What is the situation with www.amazon.ca, and amazon.co.uk? It seems to me that some numbers are global, and some are (or have been) restricted to one site's sales. Next, if a huge publishing house buys just three copies of CMOS16 (and its editors may all normally use CMOS online, as I do), this hardly reflects on the weight it should be accorded. They probably don't even buy from Amazon: Chicago sells these things directly, as one option. Same of course for APA16, which we have seen ranked at an astonishing 19: that's 19th among all Amazon's book sales. And that's only edition 16, whereas earlier editions have also been separately listed for CMOS above on this page. In fact, outside the "older" humanities APA has an enormous following, well beyond the academic discipline of Psychology and well beyond its supposedly American constituency. As we know, it supports en dashes in the "oppositional" senses quite generally:

[...] used between words of equal weight in a compound adjective (e.g. Chicago–London flight). (APA16, p. 97)

Note the wording: "in a compound adjective", and note their example of a compound adjective. APA – quaintly, as an MLA-like relic of earlier times – immediately follows with this:

Type as an an dash or, if an en dash is not available on your keyboard, as a single hyphen. In either case, use no space before or after. (loc. cit.)

Now, every journal that publishes according to APA would demand an en dash in submissions; or at the very least they would convert that hyphen to an en dash. The hyphen is a legacy of days when authors left all that to typesetters. We can't do that here; but we can acknowledge that many editors will just use a hyphen and expect other editors to fix it. Or they'll be blissful unaware of all this tumult (lucky devils). The usage in question is one normal, international standard; and CMOS is indeed unusual in stickling against the en dash in just some oppositional cases. It even uses the en dash, copiously, in ways we might interpret as proscribed by CMOS itself.

2. "American" and "British" usage

Jeff, you say above:

I′m somewhat baffled by "Chicago versus American/British/Australian usage"; unless I′ve really missed something, Chicago style is American usage, if one is to judge by sales and citations in other style guides.

I hope that is answered now! CMOS is one predominantly American style guide. It is wildly outsold, and out-used, by APA. And then there is MAU, and MWM, and so on. But I would emphasise this: just as CMOS and APA have considerable respect beyond America for their competing prescriptions, so do Oxford and Cambridge beyond Britain, for their relatively synchronised recommendations. Indeed, it is wrong to see either of these as purely British publishers (Oxford especially has a huge American production; so does Penguin, so do many other British houses). Next, it is absurd to suppose that Oxford style guides intend to express Oxford house style. It would be just as absurd to say that CMOS is Chicago's own internal record of its practice, and nothing more. Publishers everywhere defer to New Hart's Rules (NHR). Even Butcher's (BCE; Cambridge) does! So do very many miscellaneous houses, like Pickering & Chatto [search for their online pdf style guide: Wikipedia just now stopped me posting the URL, as spam-suspect!] and innumerable others. And Oxford is not constrained to follow NHR. It is adaptable to circumstances and readerships, with the sophistication and wisdom we would expect.

It is artificial to speak of any one style guide's corpus of practice as equivalent to "American usage" or to "British usage" (or "Australian", etc.). Dick is right when he opposes CMOS style to "American/British/Australian usage". If there is an international consensus out there, it is heavily in favour of the sorts of uses MOS currently supports for en dash (though the spaced version apt for long-form date ranges is less commonly accepted, and I agree that it needs limiting). CMOS is an outlier, and its own usage is against its own odd mosaic of guidelines. That's for en dash; I could say more about apostrophes and ellipsis points, but that's for another time.

Yes, sources (even otherwise "reliable" sources) do all kinds of things. We are interested in best practice, not the churning chaos that it is easy to extract using Googlebooks. Why mimic disorder and inconsistency, when there are guidelines to emulate, adapt, and enshrine for Wikipedia? MOS should set up a standard, not a register of the accumulated failings of others.

Last on this point, it would be quite wrong to think that CGEL (the most influential and prestigious grammar of English in current use) is British. It is quite explicitly international; one editor is Australian and British in his career, the other British and American. Contributors are from all over the anglophone world. It describes widespread use of WP-type en dashes, and uses them copiously, if not consistently (as I would pedantically charge the editors), throughout its 1800+ pages.

3. Print guides for Wikipedia?

Certainly we have to respect and study style guides aimed at print. There is nothing else available, except Yahoo's amateurish effort. Well, good luck to them: but we're streets ahead of them. (Incidentally, their print version is at odds with their online version in its implementation of styling. Not a good look.) And CMOS16, for all its vaunted embrace of the online world, misses the mark almost entirely: inconsistencies with terms, core issues ignored, total oldthink.

Equally certainly, we have to adapt what we find in those print style guides. This is not OR (the spectre of original research, which only applies to our articles in any case). It is a practical inevitability; and it is damn hard to do. So far MOS has performed heroically, achieving fine results against formidable opposition. I repeat what I have said elsewhere: there is no more comprehensive, more usable, more articulated set of guidelines for collaborative online writing to be found anywhere. If any editor knows of one, let's see it please.

What we are going through right now is a small but important chapter in the development of MOS: one of the more difficult chapters, sure. But among calls for adjustment we have seen strong support for the MOS guidelines as they have evolved so far, and strong support for the overall role of our community's home-grown Manual of Style. Let's maintain it, weed it of uncertainties and excess, and let the Project continue to benefit.

Noetica^Tea? 07:27, 6 June 2011 (UTC)

1. Style guides. I share Noetica’s amazement at the APA ranking, which seems to vary between the high teens and low forties, and has done so for several years. I also agree that Amazon sales are far from dispositive, for many of the same reasons stated. I don’t have enough information to speak with authority on APA’s following outside the humanities, but have a hard time putting it on overall par with CMOS, simply because it’s far less comprehensive. APA indeed seem to support most of the dash usage in the current MOS, but the coverage is so cursory that I don’t think it alone suffices; for example, APA clearly have at least two uses in the manual that aren’t discussed. I again remind that Chicago specifically disavow only one of our current uses, and do so because they don’t think most readers will notice. And they may be right; Mencken may have overestimated the intelligence of the American public. To the extent that an en dash is relied upon to make a distinction, Chicago may have a point—this would be akin to relying on red and green for critical distinctions in disregard of the fact that some people are color blind. Excluding such situations , I would again ask “What’s the problem with using en dash if some will appreciate the distinction while others will just see a hyphen?” It seems to me that using a dash accommodates both groups while using a hyphen accommodates only the latter. And Chicago somewhat contradict themselves with their long-standing example post–Civil War era; readers who see only a hyphen will see the war era that followed the time we were civil (which would seem to me to include most wars . . .). Again, I nearly always start with CMOS, but that doesn′t mean that I necessarily follow its every recommendation.

Though I happen to agree with most of the current MOS, I would question the assertion of “one normal, international standard”. Though it may surprise some here, PMAnderson may well be right in claiming the en dash is far from predominant in actual use, especially in American publishing. That isn′t necessarily controlling here (see my allusion to Mencken above); again, in most cases using it at worst does no harm.

I agree that “my guide is better than yours” is a silly argument (unless yours is from the Hillbilly Gazette). The suitability of the major guides is probably determined more than anything by subject; in the physical sciences, for example, nearly all the general guides fall far short (I don’t have CSE, so I can’t comment).

2. “American” and “British” usage. I stand by my statement, and would need considerable evidence to term CMOS as the outlier in American usage.

In the US, the “top” guides would seem to include APA, CMOS, Turabian (if you count it as separate), MLA, and still for some, Words into Type (which suffers because the last edition was in 1974). Other works, such as S&W, GMAU, and even The King’s English and MEUG are still commonly used, though they really aren’t style guides in the sense of the others (GMAU just happens to cover dashes). I′ve long liked MWM because it covers many relevant topics succinctly and clearly, but for some reason it never has attained the stature of some of the others. I assure everyone that I did not acquire its predecessor 25 years ago just so that I could torment PMAnderson today. Butcher and OSM are cited in CMOS, but both works, as well as NHR, are generally not easy to find outside of a university library.

I have little experience writing for other than American audiences, but certainly NHR and Butcher have much wider following in the UK and elsewhere than in the US, though as Noetica mentioned, CMOS and S&W seem to carry considerable weight as well.

I think most writers on all sides of the lake who are familiar with such things hold most of guides above in high regard.

That said, I think much more is being made of “AmE” vs “BrE” in this context than is indicated. The greatest difference I can find is the general British (aside from OUP) preference for spaced en rules rather than unspaced em rules for parenthetical dashes.

3. Applicability of print guides to “online” media. I don’t see what alternative we have as a starting point, save perhaps something extracted from a region in which the sunny 16 rule is inapplicable. I largely share Noetica’s assessment of the Yahoo! guide, which seems to have been developed by a bunch of, well, you get the point. For some reason, it is periodically asserted that guides for print media are unsuitable for “online” or “new” media; if that is indeed the case, I’d like a reasonable explanation of why. There are a few obvious differences, such as the lack of pagination, but I’m still awaiting answers on most of the others. I don’t think simple observation of prevalent practice is necessarily the answer. If we were to look at the majority of Web pages, often from quality sources, we would get the impression that one can safely ignore almost every tenet of professional publishing: margins, line measure, typography, spelling, and in many cases, content. There’s no problem with any of this as long as no one needs to read the material. I hope we aspire to something better. JeffConrad (talk) 03:37, 7 June 2011 (UTC)

Speaking of really awful guides, I have a fun one: Handbook for Writers by Millard (can't find a copyright date). It doesn't mention en dashes at all, but section 35e says "Use a hyphen to express inclusive dates and pages". Then it gives three examples, each of which clearly uses an en dash! Apparently, the instructions to writers, even here, do not survive a pass by the standard editor/typographer. So yes, my guide is worse than yours. Dicklyon (talk) 04:43, 7 June 2011 (UTC)

Except you don't really want to argue that the worst guides are the self-contradictory ones. Art LaPella (talk) 06:03, 7 June 2011 (UTC)

Wasn′t Millard the copy editor for the Hillbilly Gazette? See Mencken above. JeffConrad (talk) 06:15, 7 June 2011 (UTC)

Boston University. And it's really not a bad book, "for use in college composition courses". She thanks her typist. Clearly en dashes were not relevant to her advice, but somehow her editor/typographer didn't get the message. Sorry I made fun of it; it's not awful, and its error does make a relevant point. Dicklyon (talk) 06:27, 7 June 2011 (UTC)

And I hardly can claim to never have made such an error. Unfortunately, few books are error free (witness even classics such as Sidney Ray’s Applied Photographic Optics, and from what I have read, the latest edition of CSE). Whenever tempted to shout “Mine’s bigger than yours”, we should remember that, except perhaps for the Gazette, most guides agree on far more than on what they disagree, and the average writer would be far better served by reference to almost any reasonable guide than by simple extraction from, well, a familiar place. JeffConrad (talk) 06:45, 7 June 2011 (UTC)

@Noetica: “Consensus is Wikipedia's fundamental model for editorial decision-making”, hence, doing things “against formidable opposition” is most definitely not something to be flaunted. ― A. di M.​_plé​^dréachtaí 14:51, 7 June 2011 (UTC)

With respect A di M, a point like the one you've just made could be judged captious, and it risks unleashing torrents of unhelpful text on the page. Perhaps that's my own fault, yes? One or two might resoundingly assert that. But let me give a brief answer.

The effort to reach and implement consensus decisions can meet formidable opposition. It can be maligned and misrepresented in every available forum, near and far. It can be subverted by those who just don't like it, and will not accept it – despite any weight of evidence, any force of argument. MOS has performed amazingly well, surviving as a valuable resource for the Project. Right now for example we are seeing general community support for its guidelines on en dashes, intersecting with "the stretchingly difficult subject of hyphenation" (Burchfield, MEUB). I would be interested in responses to my challenge: find a better resource, for collaborative web editing. If you do find one, ask how it might succeed against odds such as our MOS contends with.

MOS is developed by reasoned argument in orderly open discussion, despite formidable opposition to that process.

Noetica^Tea? 23:10, 7 June 2011 (UTC)

What's best for Wikipedia – Break

So, yea, slight reboot here. One problem that we have is that the issues involved here have devolved into seemingly interminable bickering. Half of that problem, from my perspective, is cause by... well, what just happened above. I respect all of you, but you all seem to take this approach where you're trying to teach rather then actually addressing the issues with how your own views (what you guys are trying to teach) can work alongside with those of others. For example, the "to stand for between" usage seems to have proponents and some detractors, and each side can state their case well enough. So, that being the case, rather then trying to one up each other with "my source is better because..., so we should follow that style", my question instead is: "Which option is best for Wikipedia, and why?" That's what we're really interested in after all, and I'm just not seeing much of that.
— V = IR (Talk • Contribs) 18:18, 6 June 2011 (UTC)

I think "here's what I think and why" is a reasonable response to the question "what is best for wikipedia?" What are you looking for here? ErikHaugen (talk | contribs) 18:33, 6 June 2011 (UTC)

Honestly, I'm not even sure. I think that I'm just hoping for a slight change in tone. A change in viewpoint, I guess. I think that's what's going on below, in the #Possibilities for rewording mid-pollsection.
— V = IR (Talk • Contribs) 18:36, 6 June 2011 (UTC)

V =, I genuinely would like to know what you mean by this: "One problem that we have is that the issues involved here have devolved into seemingly interminable bickering. Half of that problem, from my perspective, is cause by... well, what just happened above." What "just happened above" that you wish to comment on, and what is your specific analysis of it, and your reply to it? I hope we can express ourselves more directly, to put any confusion behind us.

Noetica^Tea? 23:10, 6 June 2011 (UTC)

Exactly what I said in the next sentence: "but you all seem to take this approach where you're trying to teach rather then actually addressing the issues with how your own views (what you guys are trying to teach) can work alongside with those of others." There's a bunch of somewhat long-winded explanations here, but there's very little discussion about actually resolving the points that are in dispute. It just seems to me that we need less "<style guide> uses this convention" and more "Wikipedia should use this convention because". ...I'm really wondering if what's going on here is that various people are padding their positions for an anticipated arbcom case, rather then honestly trying to resolve the dispute at hand, to be blunt.
— V = IR (Talk • Contribs) 00:49, 7 June 2011 (UTC)

Thanks for explaining, V. Please assume good faith: in the case of participants above there is no reason to assume otherwise. Please do not characterise as "long-winded" what may instead be a necessarily complex presentation of complex issues. I certainly agree that brevity helps; but sometimes it is not achievable. You invited long and complex discussion when you initiated a section headed "What's best for Wikipedia", and you got it. Speaking for myself, I first addressed two themes (methodically numbered) partly to address some matters on which I deliberately held off from commenting on earlier, but mainly to arrive logically at a third: how what is best for Wikipedia is a thoughtful and unique adaptation of existing style advice. This cannot be a mere blind following, since existing advice is patently inadequate for Wikipedia's unprecedented scope, and its totally new environment. As for teaching, there is a place for that here too. Yes, punctuation is common. But the accurate study of it is specialised and multifaceted, and it is changing rapidly. We all should be ready to learn. Noetica^Tea? 01:35, 7 June 2011 (UTC)

Whoa, hang on, I have complete confidence that yourself and others here are participating here in good faith (yes, I think that even PMA is here with good faith intentions... *wink*). What I'm suggesting though is that the effors that yourself and otehrs are expending in attempts to explain the back-story to your reasoning is somewhat misplaced for the current audience. It seems to me that all of us here are well versed in the styles themselves, at this point. I just think that it's time to move on from enumerating the way that others take care of these issues and refocus on what we can and should recommend for ourselves. Far from a "mere blind following", I think that we're reaching a point of being "over-educated" here. Besides, no matter what we decide here, it's not going to be written in stone. This is a Wiki, after all! The point about things changing rapidly here should motivate us to make a solid set of recommendations rather then scare us from making any recommendations.
— V = IR (Talk • Contribs) 02:04, 7 June 2011 (UTC)

Thanks again, V = IR (Ohms law). We all have to be patient, as each editor strives to keep on track with the issues here (ill-defined as they are, truly). We could comment back and forth on each other's choices in doing so; but that too would generate noise. Let's continue to pursue the signal with diligence, respecting different understandings of what that is exactly. In particular, let's stop this now, OK? There's more focused work to do elsewhere on this page, and on its parent page. Noetica^Tea? 02:15, 7 June 2011 (UTC)

That in most cases making a distinction is not important

[I "proactively" moved this section from the voting page to this discussion page, in accord with Casliber's direction. Noetica^Tea? 13:14, 5 June 2011 (UTC)]

Deplorable - almost as deplorable as using a word in scare quotes - or pro-active in any environs. .Septentrionalis PMAnderson 21:51, 5 June 2011 (UTC)

That is, this is all a matter of Big-endians v. Little-endians still.

• Agree Collect (talk) 12:26, 5 June 2011 (UTC)
• Agree. Even the regulars are often unaware of their own rules. Art LaPella (talk) 16:30, 5 June 2011 (UTC)

I don't get it (and I'm intimately familiar with endianness). Where was this "proactively moved" from, exactly?
— V = IR (Talk • Contribs) 00:24, 6 June 2011 (UTC)

V =, see the yellow header at the top of the voting page. The topics to be voted on were determined over many days. Casliber then extracted, highlighted, and clarified them. We were invited to keep things orderly:

Editors are strongly advised to use the daughter page [= the page we are now on] for ongoing debate, and that editors clerking this page have the ability to proactively transfer material to that page.

I moved to here the discussion of a new issue – added well after the well-advertised deadline, confusing to many participants, a belittling of our whole agenda. Doing so, I cited Casliber's word: "proactively". Another editor reverted the move, commenting unfavourably on it. Questions that arise for me:

1. Who, if not us, is "clerking" these pages?
2. Do we take orderly process seriously, including deadlines, or not?
3. Do we editors act to maintain order, or not?
4. Are behavioural norms to be enforced here, or not? (See original ArbCom order.)

Noetica^Tea? 00:53, 6 June 2011 (UTC)

Ah, got it, this is just another round of seeing who's got the biggest penis. Thanks. <ignore on>
— V = IR (Talk • Contribs) 00:59, 6 June 2011 (UTC)

Heh heh. Nice take on big- and little-endianness. ☺ Noetica^Tea? 01:07, 6 June 2011 (UTC)

• Disagree, at least partially. In many cases, I’d probably concede that most people would not notice the distinction. But I’ll ask yet again: what is the problem with making the distinction for those who would appreciate it? JeffConrad (talk) 05:27, 6 June 2011 (UTC)

Okay, on pages where we really need to maintain focus and direction (especially where tempers have been frayed), it is prudent to have uninvolved people acting as clerks, as removing comments can be confrontational at the best of times. Hence me leaving it up to me. I have requested others to help clerk these two pages but so far no-one else has put up their hand (I am having slow loading times and haven't checked the history so apologies if someone has joined in). This page is for discussion, the other needs to be as clear and focussed as possible to measure consensus. I can't count the number of times I have seen debates buried in slabs of text which (intentionally or unintentionally) serve as filibusters. The last issue is actually a good one - namely the strength of MOS as policy, but we can't open that up here. A separate RfC I think is due. Casliber (talk · contribs) 01:34, 6 June 2011 (UTC)

Ok; pardon. I thought the point was to make it clear that editors don't have to figure all this out and understand the MOS before editing. This had come up before; people were expressing concern that it was hard to type dashes, etc, so I thought the point of it might be to make it clear that nobody would ever be pressured in any way to figure out how to enter a dash. But considering your interpretation, yeah, it would be nice to have that discussion as well. I would hate to have this same debate over and over as folks try to make exceptions based on taste. ErikHaugen (talk | contribs) 01:43, 6 June 2011 (UTC)

Thanks Cas – for all that you are doing here. There is much to try one's patience! So we should take it that you are the one to move material off the main page? You will understand: I sought only to maintain the order that editors of good will want here, and that you have mandated from the start. Will you please modify the yellow header to make this clear?

I request that the cryptically worded new section be moved again to here, since as you point out it is off-topic. (What is on-topic was settled in an orderly way, over many days.)

At the very least, that whole comment regarding your use of "proactively" (and the misrepresentation of my citing it, and my "deplorable" action!) is offensive and a distraction. Will you please deal "proactively" and decisively with it?

Noetica^Tea? 01:50, 6 June 2011 (UTC)

Very well Cas, thanks for putting both my explanation and the response to it at the start of this section, where they more naturally belong now. Since they are now here, I issue this clarification:

1. I acted in good faith, according to the instructions at the top of the voting page, to keep order. It is best if those instructions be up-to-date and unequivocal, since we are in difficult uncharted waters. As things stand, I believe it is uncertain what is acceptable on the voting page and what is ruled out by the agenda that was set in the initial stage of the discussion.

2. My use of the word "proactively" was a direct quotation from Casliber's instructions. I think it is unhelpful that a comment at the start of this section misrepresents that quotation, and the orderly move that I made. I am content that the comment remains on record in the history; and I do not currently intend to take the matter further.

3. Procedure and civility will be crucial, as we know all too well from what led up to all this. I make no apology for caring about such things, and I urge others to be vigilant also. Just now, though, I suggest we shift the focus back onto content. Let's hope it is not moved from there too often.

Noetica^Tea? 03:02, 6 June 2011 (UTC)

Noetica's claim that these were ever "selected as topics to be voted on" is false, unless Casliber's sudden imposition of a poll is a selection. The suppression of this topic (which is, I think, most editors' opinion of this whole page) is most unfortunate. I will consider what should be done. Septentrionalis PMAnderson 22:43, 6 June 2011 (UTC)

What exactly do you think "this topic" is? Thanks, ErikHaugen (talk | contribs) 23:05, 6 June 2011 (UTC)

It may be that some editor will wonder this in good faith: therefore: Whether "in most cases making a distinction is not important".

Differentiating between dashes and hyphens is not particularly important; and it is a pity that those who make it important are doing so in order to make Wikipedia marginally less readable. Septentrionalis PMAnderson 23:16, 6 June 2011 (UTC)

I resisted the urge to !vote here until it was (temporarily) moved to this discussion page. As Casliber put it later, "The last issue is actually a good one - namely the strength of MOS as policy, but we can't open that up here. A separate RfC I think is due." Art LaPella (talk) 00:07, 7 June 2011 (UTC)

Possibilities for rewording mid-poll

Given the discussion on the polling page that seems to indicate that some of the conditions given for the use of en-dashes (though supported in principle) are not worded in the most fortuitous way, what are the chances of making changes to the wording without having to start the poll all over again?--Kotniski (talk) 07:29, 6 June 2011 (UTC)

By way of specificity, I've had a go at recasting the section (hopefully without changing what it was trying to say). This is it (added text and my comments are in italics):

En dashes (–, also codable as &ndash;) have several distinct roles.

1. To stand for to or through in ranges (pp. 211–19, 64–75%, the 1939–45 war). Do not combine dashes with prepositionsRanges expressed using prepositions (from 450 to 500 people or between 450 and 500 people) should not use dashes (, not from 450–500 people or between 450–500 people). Do not use dashes in ranges thatNumber ranges must be spelled out if they involve a negative value, or where they might be misconstrued as a subtraction (−10 to 10, not −10–10).[Example of where they might be construed as a subtraction??]
2. Between numbers with the sense of to, in sports results and similarTo stand for to or versus (male–female ratio, a 4–3 win, the committee voted 5–4 in favourLincoln–Douglas debate, France–Germany border).
3. In compound modifiers (see "Hyphens" above) where the parts of the compound are To stand for and between independent and equal elements (diode–transistor logic, Michelson–Morley experiment, [insert here some/all of the deleted examples from the previous point]). An en dash is not used for a hyphenated personal name (Lennard-Jones potential, named after John Lennard-Jones), nor a hyphenated place name (Guinea-Bissau), nor with an element that lacks lexical independence (the prefix Sino- in Sino-Japanese trade).
4. To separate items in a list—for example, in articles about music albums, en dashes are used between track titles and durations, and between musicians and their instruments. In this role, en dashes are always spaced.[I would put this point at the end of the list, or certainly move it down under the next one.]
5. In place of a hyphen in compounds whose elements themselves contain hyphens or spaces (the anti-conscription–pro-conscription debate) and when joining a prefix to prefixing an element containing a space (pre–World War II technologies, ex–prime minister) – but usually not when joining a prefix to prefixing an element containing a hyphen (non-government-owned corporations, semi-labor-intensive industries). However, recasting the phrase (the conscription debate, technologies prior to World War II) may be better style than compounding.
6. As a stylistic alternative to em dashes (see below).

Is this at all helpful at this stage? --Kotniski (talk) 08:41, 6 June 2011 (UTC)

(after edit conflict) I would sure hate to even hint at establishing a precedent for voting on the wording of every change, which would be akin to assigning every editor a character on the keyboard and giving the directive to type away. It seems to me that the greatest objections have been to

1. “to stand for” in a situation where this phrase cannot directly substitute for an en dash. Perhaps we could try something to the effect of “in the sense of whatever” (and in some cases there may be more than one whatever)
2. Possibly splitting hairs with some of the different uses. I may have gone overboard in the breakout, but the objective was to avoid “partial disagreement” as much as possible. I′m all for eliminating artificial distinctions if we determine that we have consensus for a range of subcategories, but I′d like to be sure that we really have it before recombining subcategories only to be shot down at the last moment (we have until 16 July, after all . . . arrgh).
3. Some of the examples (my male–female ratio is not even being praised by faint damns . . .). I don′t have a direct suggestion here, except that the wording and the example should agree and make the meaning is clear. As did Chicago, we probably need to accept that it′s simply impossible to cover every case, though we should try to address the obvious ones.

JeffConrad (talk) 11:12, 6 June 2011 (UTC) (Jeff’s actual signature added after section split)

The spacing issue

The issue of spacing en dashes may not be as simple as I thought: “when there is a space within either one or both of the items”. There seem to be several different cases, and it seems that we may agree on some but not all. Consider

• New York – Sydney flight or New York—Sidney flight?
• New Zealand – South Africa grand final or New Zealand–South Africa grand final?
• San Luis Obispo – San Juan Capistrano trip or San Luis Obispo–San Juan Capistrano trip (or neither)?
• 3 – 6 June 1888 or 3–6 June 1888?
• June – August 1888 or June–August 1888?
• June 3, 1888 – August 18, 1940 or June 3, 1888–August 18, 1940?
• 3 June 1888 – 18 August 1940 or 3 June 1888–18 August 1940?

There seems to be general agreement that the last example is potentially confusing with the dash closed up (some of us would simply prefer to or through), but where we stand on the others is less clear. American guides and OSM seem to call for closed up use in all cases (though I haven’t found the last example in any that I’ve read, and I assume there is some limit to the number of words in a compound that a dash, spaced or unspaced, can reasonably connect). Though I′ve only glanced at a couple, the approach of British guides other than OSM seems different. Butcher states, “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” [emphasis added]]; the operative question would then seem “When would closed-up use confusingly imply the wrong closeness of the relationship?”

I think the exact wording and perhaps even the examples are details that should not affect most editors’ votes. But 6b seems to be a weightier issue; with everything lumped together, we seem to agree to disagree, so I think we need to see if we can agree on some of the individual cases, though not necessarily as many as I show above (hopefully, I haven’t overlooked any). Quite honestly, the thought of this discussion continuing through 16 July boggles the mind. Anything we can do to at least see where we stand long before that seems a good idea. JeffConrad (talk) 09:22, 6 June 2011 (UTC)

For me, there's a distinction between the cases of "compound modifiers" (compounds used like adjectives, before the noun) and other cases (mainly ranges). Despite many years on Wikipedia it still looks wrong to me when I see a spaced dash used in a compound modifier, as in South Korea – Thailand relations (which looks like it means something like "South Korea, and in particular the Thailand relations"). In other cases, like date ranges, I don't have any objection (though I certainly wouldn't space the dash in cases like 3–6 June or June–August 1888; here the space doesn't come within the items being joined).--Kotniski (talk) 10:37, 6 June 2011 (UTC)

Yes, the latter would simply be incorrect. No-one recommends them. — kwami (talk) 23:50, 6 June 2011 (UTC)

And based on the comments, I suspect we′re not alone. The closed-up use with the “international” form of the full-date range is a mess, and I would probably avoid the dash, spaced or unspaced, with the American form as well. With alphabetics (probably just compounds) it seems a bit different. Though it’s conceivable that closed-up use might imply the closest relationship between the words adjacent to the dash, it seems to me that we’ve already addressed this by using a dash rather than a hyphen. If we have a problem with London–New York flight, we certainly have a problem with post–Civil War era, do we not? JeffConrad (talk) 11:12, 6 June 2011 (UTC)

Jeff, those are not parallel. A London–York flight would have a dash, whereas post-war era would not. A problem with one does not imply a problem with the other. — kwami (talk) 23:50, 6 June 2011 (UTC)

I don’t quite see the relevance here; at issue is whether mistaken close association would lead to confusion. Though London-New York flight would be confusing, I doubt many people would take London–New York flight as meaning the flight from London-New to York; we’ve already addressed the question of mis-association by using the en dash to indicate “London-(New York)” rather than “(London New)-York”, so the space should not be needed. That almost every style guide I’ve read (previously cited) show this usage closed up argues strongly against mandating or even encouraging it. JeffConrad (talk) 01:20, 7 June 2011 (UTC)

Here′s a good example of how it’s easy to get carried away with trying to join too many open compounds with an en dash. And offhand, I’d say Chicago staff still misread “medium-charge density”. Sometimes, almost any solution is ugly, but some are additionally unclear. JeffConrad (talk) 06:58, 7 June 2011 (UTC)

Specific rewording suggestions

Specific comments on Kotniski′s suggestions:

(1) I′m fine with the proscriptions. It seems to me that −10–10 is the example of possibly implied subtraction (I had to read the code to distinguish the dash from the minus); what may be needed is the example of a range with a minus (−20–−10). The latter example is so ridiculous that it’s tough to imagine anyone using it, but I would not be surprised if some editor somewhere might sometime prove me wrong.

(2) OK, though I think 5–4 decision is just as good as well as more succinct, and it’s one that commonly appears in the news, usually with a hyphen . . . The simpler version would also avoid an arguably gratuitous British spelling. I concede that the proposed example might illustrate a slightly different sense.

(3) “To stand for” clearly isn’t the right wording, but I’m not sure the proposal is sufficiently clear. Though APA use this wording, many others use wording “in the sense of to, and, or between”. Perhaps some combination thereof would work. I’m not sure we need separate examples of each usage, but I think merger at this point is premature because we seem to have agreement (if grudging) on some senses but not others. I think “nor a hyphenated place name” is missing with or in after nor. I′ve always thought “an element that lacks lexical independence” is an unnecessarily hifalutin way of saying “an element that cannot stand alone”.

(4) I agree with moving this to the end of the list, but without better explanation, and at least one example (such as from Kind of Blue), it requires the reader to guess at what’s meant.

(5) Again, we have partial disagreement, so I think recombination is premature. There may be an additional issue: is there a limit to the number of words in a compound (e.g., {xt|pre-world War II technologies}}; was this the reason for WFC’s objection?). The spaced en dash (in the current version) is disconcerting in combination with the unspaced em dashes elsewhere; if the intent is to illustrate both, I find the effect overly cute.

Again, I think the wording is largely a detail (if a significant one) that probably won’t affect most votes. More important, perhaps, is that it seems a few potential issues may have been overlooked (I did the breakout, so I guess I overlooked them), such as use with ratio and use of a dash rather than a solidus to indicate alternatives or something similar. We also need to resolve spacing in various situations. JeffConrad (talk) 10:17, 6 June 2011 (UTC)

Rewrite, not reword

I don't think we're going to learn anything more by trying to reword what we're voting on. Let's start a rewrite instead. It looks to me like the 6 points collapse to 4, with numbers 4 and 6 on dash alternatives merging, and numbers 1, 2 and 3 collapsing into one on numbers and dates and one on compounded words, or something like that. In the process, we need to get especially clear on how to resolve the theories that were advanced in getting us into this mess, like whether the en dash applies between equal forms other that nouns (French–German border), and in forms where the parallel relationship is not obviously one of the set of (and, to, versus, between) or some such (love–hate relationship); and include examples such as these from actual guides, or better ones of our own choosing. Dicklyon (talk) 15:16, 6 June 2011 (UTC)

The essential point is that en dashes link independent elements. Page and date ranges are a special case of that, probably best kept separate, since they're not generally attributive. There is no indication that part of speech has anything to do with it. The preposition used when saying it out loud is also rather irrelevant; no matter which ones we use in the description, there will be cases where they're used for non-independent elements, and where independent elements would be best worded with some other prep. — kwami (talk) 23:46, 6 June 2011 (UTC)

The essential point is that, in normal usage, ranging between a small majority to an overwhelming majority, independent elements are linked by a hyphen. A dash is an acceptable variant in those few cases in which a dash has a meaning distinct from that which a hyphen would have; these are also the cases in which a dash is a significant minority of usage.

Those who wish to change English should do so elsewhere and among others. Septentrionalis PMAnderson 00:45, 7 June 2011 (UTC)

I really wish that you'd quit with the Defender of the English Language polemic. It's tiring and does nothing to help resolve any of the issues here. Actually, it does nothing except to polarize sides, since it understandably causes defensive reactions. I understand that I'm making myself a target here, which I've been trying to avoid, but I'm just really getting tired of seeing this same refrain over and over again. You don't control the English language any more than myself or anyone else here does, you know.
— V = IR (Talk • Contribs) 01:01, 7 June 2011 (UTC)

I decry only those who wish to ignore usage entirely, in favor of one or another version of "correct typography". They would make the encyclopedia less readable if MOS had great power; as it is, they merely make MOS look silly. Septentrionalis PMAnderson 17:00, 7 June 2011 (UTC)

I guess I′m having a hard time seeing how what many are suggesting here would make the encyclopedia less readable, especially for readers who supposedly would not notice the difference. Similar for making the MOS look silly. The current MOS certainly isn’t the only way to do it, and it doesn’t find universal support (almost nothing does), but most of it does find considerable support in many highly regarded guides. JeffConrad (talk) 23:48, 7 June 2011 (UTC)

When the black UN helicopters that UOP have hired descend in the middle of the night to destroy our language, we may note that PMAnderson warned us repeatedly, but we paid no heed until it was too late. But I honestly won’t be staying up all night with assault rifle in hand ready to repel the invasion.

I think Ohms law and I probably speak for many others who are more than weary of this pointless polarizing polemic, which serves mainly to set others against almost any position PMAnderson might espouse, regardless of its merits. There′s certainly room for disagreement, but let’s try to focus on the issues without needlessly stirring the pot. We’ll accomplish far more without the unnecessary angst. JeffConrad (talk) 01:37, 7 June 2011 (UTC)

I tend to agree that a rewrite is preferable to a few tweaks, but we still first need to find what agreement in principle that we can or the wording won’t matter. I would probably assemble reasonably comprehensive examples on which we can agree and see how many discrete cases they suggest. As I mentioned, I think “between independent elements of a compound modifier” is too abstract; perhaps we could use it in combination with some mention of “sense”, using whatever words may be appropriate.

I am still reluctant to combine (4) and (6) because I still don’t completely understand (4)—it seems a use we have developed here. If indeed (4) and (6) can be combined, an unspaced em dash could be used as an alternative, and I’m not sure we mean that. Several of us have suggested that a musician and his or her instrument could be linked using a comma, colon, or perhaps an em dash. But I don’t really know whether (4) permits, proscribes, or ignores these alternatives, so we need to make clear what we mean. It’s been suggested that the application is not limited to music albums; if indeed there are other specific applications we should describe them.

In any event, we need to find reasonable agreement in concept before tweaking the wording. I’ve been involved in far too many efforts that got this backwards, ultimately making the task far more difficult. JeffConrad (talk) 02:00, 7 June 2011 (UTC)

I'm not well versed in the specifics behind (4), but I wanted to offer it some support here because it's the type of "rule" that I think we should encourage within our MoS. I think that we can all agree that Wikipedia covers a huge breadth of subject matter, which makes the creation of completely abstract guidelines basically impossible in my opinion. Allowing for specific exceptions makes our jobs much more difficult, but I think that it's the only way that the MoS can possibly remain relevant to our editor-authors. I'm not saying that it shouldn't be combined with (6) in some fashion, and I'm not saying that it shouldn't be reworded or anything, but I don't think that it should be completely removed.
— V = IR (Talk • Contribs) 02:35, 7 June 2011 (UTC)

I’m not suggesting that we get rid of (4), but simply that we adequately explain what it is, and if appropriate, what it prescribes, permits, and proscribes. I′m still not well versed in the specifics, despite looking for and finding an example (Kind of Blue), which I linked to the discussion and the breakout. JeffConrad (talk) 07:04, 7 June 2011 (UTC)

(4) appears to be a relic of a bygone dispute. By finding when it was written in, we would probably find that it was a dispute between two insistances, one that all record articles be written Pretty Song – 4:32 and the other that we should always write Pretty Song (4:32) or some other alternative. Since neither contender seems to be around, we could probably omit it without loss. Septentrionalis PMAnderson 17:00, 7 June 2011 (UTC)

If that’s indeed the case, I tend to agree. I slightly prefer the latter form, but think either is acceptable. Similarly, I think any of Miles Davis: trumpet and leader; Miles Davis, trumpet and leader; Miles Davis – trumpet and leader; and possibly Miles Davis—trumpet and leader would be acceptable for musicians and instruments (the order could also be reversed). If this material is retained, it might be better in a topic-specific section or article. In any event, “to separate items in a list” is misleading because, with normal terminology, the list items are the tracks and the musicians (or instruments), and the things separated are elements within those list items. JeffConrad (talk) 23:59, 7 June 2011 (UTC)

Now see, here's the crux of the matter. I could go either way with this really (and I'm all but certain that PMA is correct about it's origins), but I think that getting rid of specific advice completely or pushing it off to topic specific areas is not the best approach... within reason. Things should be generalized as much as possible (we don't want massive lists of examples and exceptions after all), but I'm not at all averse to allowing for, and specifying, exceptions. I believe that one of the reasons that we're here is because there's been a reticence to allow specific advice, which is somewhat understandable because we don't want a Tome of Rules (I don't think), but that stance has the side effect of leaving what advice we do give as being wishy-washy. So, the question to me is, how to we approach this so that we're toeing an indistinct line between being overly-prescriptive and overly-descriptive. It's a feel thing...
— V = IR (Talk • Contribs) 01:20, 8 June 2011 (UTC)

I certainly agree with generalizing, but it seems here we offer very specific guidance in a section that’s otherwise quite general. If we retain this, we need to at least correct the wording, and probably make clear that several other perfectly reasonable alternatives are not proscribed. JeffConrad (talk) 03:37, 8 June 2011 (UTC)

Possibilities for rewording mid-poll - Phrasing

I just wanted to mention that I like the idea of changing "to stand for" to be "in the sense of". That seems less... I'm not sure how to describe it, exactly. "authoritative"? Actually, "wikt:imperative" is probably most accurate, where "imperative" indicates something that much be done. I realize that we're writing the MoS here, and the "rules" should be as clear as possible, but... I'm uncomfortable without leaving that little bit of wiggle room, you know?
— V = IR (Talk • Contribs) 18:34, 6 June 2011 (UTC)

Well, by being less literal, it might counter the “direct substitution” objection. I′m not sure we would even need to associate specific senses with specific uses; in combination with “joining independent terms in compounds” or similar, this wording might serve to better explain what we’re trying to accomplish, which I’m not sure any of the guides we’ve discussed adequately cover. As for absolutism, I think it’s nearly impossible for any guide to be completely prescriptive, because it’s impossible to cover (or even envision) every possible situation. Sometimes, it necessarily comes down to the editor’s judgment, as most “authoritative” guides (note the difference from “authoritarian”) guides readily indicate. JeffConrad (talk) 07:16, 7 June 2011 (UTC)

I was in favour of covering most of these by saying something like "joining independent and equal elements" (as in my suggested rewording above), without trying to analyse what the dash "stands for" or even "has the sense of"; but then I thought of yes-no question – would anyone want to write that with a dash? If not, why not? Is it because the sense is "or" rather than "and", or is it simply usage? --Kotniski (talk) 08:58, 7 June 2011 (UTC)

Perhaps because it's essentially synonymous with affirmation question, so 'yes' and 'no' are not very independent as elements. But in the end I suspect that it is simple familiarity, and the lack of any real possibility for confusion. "Yes/no question" with a virgule is quite common, and where the author is being unusually precise, as in Tsujimura An Introduction to Japanese linguistics or Schachter & Otanes Tagalog reference grammar, you do see "yes–no question" with a dash. — kwami (talk) 11:08, 7 June 2011 (UTC)

(ec) I don't see why they're any less independent than are Lincoln and Douglas in "Lincoln–Douglas Debates". By independent I thought we meant capable of existing alone, i.e. unlike the prefix "Anglo" in "Anglo-American relations".--Kotniski (talk) 11:17, 7 June 2011 (UTC)

No no, "independent" in the sence of disjunction/conjunction. Guinea-Bissau would not take a dash because, while capable of standing alone, Guinea and Bissau are not contrasting elements like yes and no or Lincoln and Douglas; rather, one is specifying the other (the Guinea of Bissau, not the Guinea of Conakry). Guinea–Bissau with a dash would suggest two separate countries, Guinea and (Guinea-)Bissau. Some people actually do dash Anglo–American relations, BTW, but it's very common to make prefixes an exception to the semantic rule. I suspect that's because there's generally little chance for ambiguity. — kwami (talk) 11:21, 7 June 2011 (UTC)

I've lost you now - you now seem to be agreeing with me that "yes" and "no" are in the same category as "Lincoln" and "Douglas" as far as independence is concerned. Anyway, back to the basic point - do you think our style guidance should imply use of a dash in (cases like) "yes-no question"?--Kotniski (talk) 11:28, 7 June 2011 (UTC)

Sorry, just my musings. My point was that in a yes–no question, there's either a yes or a no, not both as there are both Lincoln and Duoglas in the debates. Thus it's more like a "non-wh question". I was trying to explain why the dash doesn't seem as important here, at least to me. But they are nonetheless independent contrastive elements, and so formally are equivalent to the debates example. — kwami (talk) 11:33, 7 June 2011 (UTC)

But with your second point, we come to an issue that has been kind of raised here, but hasn't been included in the poll (and perhaps ought to be) - to what extent are we prepared to allow exceptions to the style rules worked out here, based on usage in particular instances as observed in sources?--Kotniski (talk) 11:24, 7 June 2011 (UTC)

IMO, because we're an encyclopedia, we should strive for precision. Thus IMO we should hyphenate high-school student even if many sources don't bother, and likewise we should dash yes–no question – though in this particular case a virgule is unambiguous as well. — kwami (talk) 11:28, 7 June 2011 (UTC)

Ah, right (though the strength of our observance of this principle should probably be put to consensus, since there is clearly some disagreement from PMA and others). Anyway, if we are supposed by be dashing things like "yes-no question", then it seems we need either a new category for dash use or a more general formulation, since we don't currently have "in the sense of or" among our suggested usages.--Kotniski (talk) 11:35, 7 June 2011 (UTC)

That's precisely why I gave the MWC10 excerpt below. I think we've gotten too hung up on coming up with the perfect prepositions and conjunctions to use as a substitute for the rule. There is no perfect correspondence, because prepositions and conjunctions have such a wide variety of use. No matter what we come up with, someone will be able to come along and say, 'but this preposition isn't listed in the MOS, so we can't use a dash here, even though it would make sense to do so' or 'but that preposition would be used in this case, so we must use a dash here, even though it makes no sense to do so'. Far better IMO to give a general guideline (elements of equal weight, independent as entities, can be reversed, etc.) and follow this with examples from all the relevant points in this discussion, without spelling out those points and arguing endlessly about the perfect wording.

Also, in the case of the yes–no example, because they're alternatives, AFAIK a virgule would work just as well, and is more common. It therefore wouldn't be a good example for us. (That's where I was headed above, but it's late at night and I didn't think it out well.)

Came across an interesting example of the difference: a Japanese–English/German/French dictionary. That is, from Japanese to English/German/French (independent and equal sides dashed, alternatives within one of those sides slashed). — kwami (talk) 11:54, 7 June 2011 (UTC)

I certainly agree with you that most of these cases should be covered by a general statement (like the "independent and equal elements" thing I used in my suggested rewording) than by trying to say what words the dash is equivalent to.--Kotniski (talk) 13:22, 7 June 2011 (UTC)

@kwami: “Many sources” is a gross understatement. High school students is spelt this way by the overwhelming majority of published books in English.[1] ― A. di M.​_plé​^dréachtaí 14:45, 7 June 2011 (UTC)

Thought I'd through this out here again. MWC10 gives a very brief treatment, but perhaps instructive. (Note they do not advocate dashes for compounded compounds and so do not need spaces for New York – Paris.) They have this under their section on hyphenation, and note for these two points that "In typeset material the longer en dash is used":

1. Serves as an equivalent of through or (up) to and including when used between indicators of range such as numbers and dates.

   pages 40–48

   the years 1980–89

2. Serves as the equivalent of to, and, or versus in indicating linkage or opposition.

   the New York–Paris flight

   the Hardy–Weinberg law

   the Lincoln–Douglas Debates

   The final score was 7–2.

This kind of conflation is what some of us have been advocating. Perhaps a general description like this, followed by a variety of examples, will be less subject to wikilawyering than trying to spell out which prepositions are equivalent to which examples. — kwami (talk) 11:03, 7 June 2011 (UTC)

What does your "MW" refer to, Kwami? There are two Merriam-Webster style guides in our list at the top (MWG: Merriam-Webster’s Guide to Punctuation and Style; and MWM: Merriam-Webster’s Manual for Writers and Editors); but there is also the Collegiate Dictionary, and other things. The Collegiate includes a brief guide to punctuation. Noetica^Tea? 11:18, 7 June 2011 (UTC)

Ah, sorry. The Collegiate, 10th ed. I thought it was a nice illustration of how compact and yet clear we can make this. — kwami (talk) 11:30, 7 June 2011 (UTC)

Thanks. If you don't mind, I'm altering your abbreviation for it, and adding "MWC" to the list at the top. The current MWC11 has essentially the same as your MWC10, but strangely omits anything like "the Hardy–Weinberg law". Noetica^Tea? 11:36, 7 June 2011 (UTC)

A random example of a non-musical article with list dashing: Eurypterus#Species. This format is quite common in our biology articles. (Em dashes, colons, and parentheses would work just as well, and are commonly used too.) — kwami (talk) 12:31, 7 June 2011 (UTC)

I think this is on the right track. I might begin with something to the effect of “independent words of equal weight in a compound adjective” (redundant?), and “often in the sense of to, and, versus, or between” (the last would cover France–Germany border). We might need to make clear what′s meant by independent and equal weight, preferably with examples showing words that are not independent (e.g., Sino-) and words that are not of equal weight (presumably, those that just take a hyphen). I say this because several experienced editors here have had to guess at the meaning; if we have to guess, what will happen with the average editor?

The entry for en dash in APA is a good example: it says only “words of equal weight in a compound adjective (e.g., Chicago–London flight)”. Though APA actually supports almost every use in the current MOS, it’s hardly obvious from this example—you need to find the other examples elsewhere in that book, and expecting an editor to go to such an effort is unrealistic.

I’d also include the example of post–Civil War era and ex–prime minister just to make clear that hyphens are confusing and unacceptable in those instances.

We might also suggest a limit to the number of words (pre-World War II technologies is probably clear, but strikes me as pushing it). Similarly, San Francisco–San Luis Obispo flight seems pushing it, though I′m not sure of the alternative, since we would use San Francisco–Santa Barbara flight, and I would find it confusing if the form varied with source and destination.

Some special cases may also need mention; for example, APA uses male-to-female transgender person, suggesting they would not use male–female transgender person. Although the sense is still of to, there’s an additional sense of transformation, which might explain the APA usage. There’s also Chicago’s example of University of Wisconsin–Madison; how would we handle this? Analogous is the common example of indicating the party and constituency of a member of the US Congress; following Chicago’s example (and arguably logic), we would use Leahy (D–VT). yet this is almost universally shown as Leahy (D-VT). Which would we use? I′m sure there are many other similar special cases, and we obviously cannot cover them all.

Because any of these uses is notational shorthand, I think it’s important to indicate what’s meant, just so an editor can paraphrase to see if a construction makes sense (e.g., New York–London flight means “the flight from New York to London” [fine], Michelson–Morley experiment means “the Michelson and Morley experiment” [fine], France–Germany border means “the border between France and Germany” [fine], but yes–no question would mean “yes or no question” [which seems less OK to me—I’d use a virgule]). Absent this ability to paraphrase, a “sanity check” becomes more difficult.

In summary, I’m suggesting we take a more expansive approach, in effect combining what’s in several different guides (such as MWC10 and APA) just to make completely clear what’s meant. The guesses that some have made in this discussion. And these are, for the most part, experienced editors who largely support the current MOS . . .

But we still first need to assess where we stand. If we can′t find agreement on the usage, the wording doesn’t matter. JeffConrad (talk) 22:30, 7 June 2011 (UTC)

• How about consolidating 2b, 2c, 2d, 3a, 3b and 3c into “In constructions like X–Y Z meaning "the Z from X to Y", "the Z between X and Y", "the Z of X and Y" or similar: examples go here”? ― A. di M.​_plé​^dréachtaí 15:32, 8 June 2011 (UTC)
  • That consolidation sounds a good way to go, though we have to be careful with the wording; the "France–Germany border" might be the border between France and Germany, but the "French–German border" is not the border between French and German (I think someone already made a point very similar to this).--Kotniski (talk) 16:30, 8 June 2011 (UTC)

  I’m not big on the use of symbols in this context (and I’m usually one to err on the side of too much math rather than too little). Ostensibly, the precision is appealing, but I think it invites the “pie are round” objection. And the precision may be a greater weakness than benefit, because in being precise, we would need a pretty large number of examples to cover the various uses (assuming we could find reasonable agreement). A big issue, again, is in trying to associate particular pronouns with particular uses; though the pronouns we′ve mentioned seem reasonable to me, they may not be the only reasonable ones, and I′m not sure it’s necessary or even possible to come up with a comprehensive list. As long as an editor has a reasonable idea of what is meant, I think he or she can select a reasonable way to express it.

  I′ve assembled a collection of usage descriptions (all American!) at User:JeffConrad/En dash that would probably cover 2b–2d, 3a–3c, and perhaps others that I overlooked in the breakout here, along with a few comments. Because the description from MWC10 is covered here, I haven’t repeated it, but would include it in ideas to consider. I don’t think any one description really nails it; I tend to like a combination of the original Merriam-Webster’s Standard American Style Manual, APA, and CMOS 15th and 16th eds. Especially interesting is Chicago’s apparent gradual metamorphosis from prescription to somewhat reluctant recognition that Mencken may have misoverestimated the intelligence of the American public. And perhaps to a lesser extent, even that of some of our friends in the Commonwealth. JeffConrad (talk) 04:32, 9 June 2011 (UTC)

  Jeff, thanks for your page of quotes and such. You asked of one "It’s also not necessarily obvious what’s meant by equal weight; would this suggest equal weight in Arab–Israeli conflict but unequal weight in Arab-Israeli spouse?" My take on it is that here "Arab–Israeli" refers to the Arab and Israeli people or governments, two equal or opposing entities, while "Arab-Israeli" is a compound descriptor of a single person, the spouse, an Israeli of Arab extraction, I suppose. It's a lot like the distinction indicated in Mexican–American War and Mexican-American culture. Exactly how to express this distinction may remain somewhat elusive, but easy to decide when the intent is known; similar to the distinction between the French–German border and the French-German culture of Alsace-Lorraine, perhaps? As Hullaballoo Wolfowitz points out of some of these usages, "requires interpretation to apply." This is a good thing, for editors willing to make interpretation based on knowledge of how en dashes are used. If no interpretation were required, the en dash would not convey any information, would it? Dicklyon (talk) 05:05, 9 June 2011 (UTC)

  Dick, we clearly have no disagreement on the meaning of “Arab-Israeli” or “Mexican-American”—we both clearly understand the obvious distinctions with “Arab–Israeli” and “Mexican–American”, and we both recognize that the distinction is significant. The issue I was trying to raise was whether the meaning of “equal weight” is clear. If we were to ask the question in San Jose on Cinco de Mayo, I’m not sure all respondents would say that Mexican was of lesser weight.

  So I think we may need more than the APA description.

  To those who would say that the distinction is unimportant or that most people would not recognize the en dash’s role in conveying it, I would say that, at worst, those who recognize the distinction will appreciate it; those who do not will have suffered no loss. So why not accommodate those who do appreciate it? JeffConrad (talk) 05:37, 9 June 2011 (UTC)

  Exactly. And that's why I prefer the guides that say "parallel", or "equal" in some structural sense other than weight, or "independent". I was just looking at my latest to arrive, the American Chemical Society's ACS Style Guide 2nd ed., which has several narrow cases as many do, but on colors recommends a hyphen: "Hyphenate combinations of color terms used as modifiers." But this is bogus, since "blue–green" or "blue-green" might be correct, depending on whether it refers to a blue versus green or a bluish green. It's not a weight difference, but an asymmetry. Other guides use this example specifically, as here in The Chemist's English, in which the distinction is expressed as between merging the two into one concept versus maintaining the two separate identities; this seems closer to the mark. It's usually not hard to decide which sense one wants, if one has a bit of experience and "ear" for punctuation, which is the point made in Bugs in Writing. Guides aren't going to save us; we'll still have to think. But we can try to distill their advice into something as good as the best, or a combination thereof. Dicklyon (talk) 06:00, 9 June 2011 (UTC)

  Just remember that neither guides nor thinking will save us, unless the sight of a word like "epileptics" makes an alarm go off, that tells us it's time to look up the rule for that word. Art LaPella (talk) 06:24, 9 June 2011 (UTC)

  Art: ¿Qué? Noetica^Tea? 08:48, 9 June 2011 (UTC)

  If you don't know or can't find the Manual's rule on epileptics, then you have made my point better than linking to it would accomplish. Does the epileptics rule accommodate those who appreciate it, while doing no harm to those who won't notice? Or is it all in all, just another brick in the wall that prevents people from understanding the rest of the Manual?

  Having said that (I've said it before, but the "does no harm" quote keeps getting repeated), obscure rules isn't the main issue on this page. Septentrionalis's preferred rules on dashes are as obscure as anyone's, [I retract that now; I discovered it isn't true] and his expectation of people studying a table full of style guides and other indications of usage is even more unrealistic than expecting people to study the half a megabyte of rules in the Manual of Style including subpages. But I'm glad to see people mentioning instruction creep occasionally. Art LaPella (talk) 20:28, 9 June 2011 (UTC)

  For the most part, I think we can rely on general principles here rather than special rules for special words. If someone uses a hyphen where we might recommend an en dash, the world as we know it is unlikely to end. Editors use hyphens all the time in number ranges (where we seem unanimous in preferring an en dash), and these cases usually eventually get corrected. Or not. My “no harm” point is that if we use an en dash where some may prefer a hyphen or see no difference (e.g., post–Civil War years vs. post-Civil War years), I just don’t see how this makes the article unreadable. JeffConrad (talk) 02:05, 10 June 2011 (UTC)

  I don't expect most people to study style guides. I would prefer, as Dicklyon appears to say below, to permit both styles; I would recommend following overwhelming usage. It is failure to do that that looks silly. Those who write the MOS should consult style guides; if there is a dispute about dashes, it can be settled by considering the sources, and by considering style guides. Septentrionalis PMAnderson 14:33, 11 June 2011 (UTC)

  While we′e considering subtle distinctions, we probably should note that APA (at 3.14) do not use a hyphen in an open compound that describes an ethnic identity, even in adjectival context (e.g., Asian American participants). APA are generally not in the business of discussing international conflicts, so I’ve not been able to find an example like Arab–Israeli conflict, but similar uses (e.g., in Table 4.5) suggest they would use an en dash. JeffConrad (talk) 02:20, 10 June 2011 (UTC)

Advertising the discussion

Has this discussion been advertised in all the relevant places? WT:MOS, WP:VP, WP:CENT and Wikipedia:Requests for comment/Wikipedia style and naming at the very least, but a watchlist notice would also be great. (I came back from a Wikibreak following a post on my user talk page, so I dunno whether that's already been done.) ― A. di M.​_plé​^dréachtaí 20:30, 8 June 2011 (UTC)

Not the last one (and possibly not the first one) - if someone could do that one, that'd be great. Casliber (talk · contribs) 21:30, 8 June 2011 (UTC)

Harry Potter and the Deathly Hallows - Part 1

I think we would (nearly) all agree that this film title needs an en dash rather than a hyphen or em dash – however it doesn't seem to fit any of the cases currently listed (either at WP:DASH or in the poll) as requiring en dashes. Have we missed something? (It seems to me to be basically the same usage as the "separate items in lists" one, though more general as it is not restricted to lists.)--Kotniski (talk) 06:53, 10 June 2011 (UTC)

I’m not sure there′s anything wrong with Harry Potter and the Deathly Hallows—Part 1 or Harry Potter and the Deathly Hallows–Part 1; US practice (except for newspapers) generally doesn’t space em dashes or en dashes. The first example is essentially using the em dash in place of a colon (S&W Rule 8, CMOS16 at 6.82, OSM at 5.10.10, NHR at 4.11.2); the second example here is possibly equivalent to Chicago’s University of Wisconsin–Madison (CMOS16 at 6.81).

Incidentally, another good example of the spaced en dash in Wikipedia is a calender day page, such as June 10. JeffConrad (talk) 09:14, 10 June 2011 (UTC)

So do you agree that at WP:MOS we currently don't say anything about this usage (except in the case of lists)? Or is it hidden away somewhere that I've missed?--Kotniski (talk) 11:22, 10 June 2011 (UTC)

It's a stylistic alternative to an em dash. The calendar page one is that, too, in the context of separating parts of a list item. It would be good to have a more comprehensive explanation of all these, but probably not focus on cases. Dicklyon (talk) 15:16, 10 June 2011 (UTC)

In that case, this set of usages seems to be missing from what we say about em dashes (which deals only with interruption within a sentence, which these uses are certainly not). Though I get the impression that for this type of use, we actually prefer the spaced en dash (don't we?) – I mean, we wouldn't write "Harry Potter and the Deathly Hallows – Part 1" with an unspaced em dash, would we? even if the sentence-interruptor used in the article were the em dash. (Presumably this is why the "items in a list" case is listed explicitly under the uses of en dash.) If this issue is controversial, perhaps it should be added to the poll (in fact some of the points in the poll can probably be closed already per SNOW, leaving us to focus on the matters that are actually seriously contested).--Kotniski (talk) 15:24, 10 June 2011 (UTC)

WP:MDASH talks about parentheticals, and then this is a "stylistic alternative to an em dash". Is this not a parenthetical? ie, we could have called it "Harry Potter and the Deathly Hallows (Part 1)". ErikHaugen (talk | contribs) 21:14, 10 June 2011 (UTC)

It’s probably a stylistic alternative to an em dash, which in turn in these instances may be a stylistic alternative to something else, such as colon; as such, it is “stronger than a comma, less formal than a colon, and more relaxed than parentheses” (S&W, Rule 8). JeffConrad (talk) 00:01, 11 June 2011 (UTC)

I don't really agree that this fits the "parenthetical" use of the em dash as described at WP:MDASH – there aren't two of them, and it doesn't represent "interruption within a sentence" which is the overriding description given. It certainly looks more like an alternative to a colon – but that's not mentioned anywhere. Can we formulate a description of this usage and add it to the poll? --Kotniski (talk) 10:26, 11 June 2011 (UTC)

The “closing” parenthetical dash of usage 1 is omitted when it'd at the end of a sentence (as is the “closing” parenthetical comma of non-defining relative clauses etc.); in some cases, a dash could be taken to be either as usage 1 or as usage 2 with little or no overall difference. I think this is one of those cases, as Harry Potter and the Deathly Hallows; Part 1 and Harry Potter and the Deathly Hallows (Part 1) would mean the same. ― A. di M.​_plé​^dréachtaí 16:47, 11 June 2011 (UTC)

But here it isn't either of those usages, as they are both listed as cases of interruption within a sentence, and here we don't have a sentence. (That's another point: how would we write it in the middle of a sentence? "...when the film Harry Potter and the Deathly Hallows – Part 1 came out..." or "...when the film Harry Potter and the Deathly Hallows – Part 1 – came out..."?) --Kotniski (talk) 17:22, 11 June 2011 (UTC)

I might use the latter, by analogy with he moved to Pasadena, California, at the age of 12 (note the comma after California), but I think not many people would do the same. ― A. di M.​_plé​^dréachtaí 17:29, 12 June 2011 (UTC)

I tend to agree with A. di M. that the use is essentially parenthetical, though I might find the semicolon confusing, especially in the middle of a sentence. It seems to me that we have a problem with rote use of either unspaced em dashes or spaced en dashes, which may argue for the way I think Chicago would do it. We might have

<beginning of sentence>—when the film Harry Potter and the Deathly Hallows–Part 1 came out—<continue>,

though I would probably prefer

<beginning of sentence>—when the film Harry Potter and the Deathly Hallows, Part 1 came out—<continue>

or

<beginning of sentence>—when the film Harry Potter and the Deathly Hallows (Part 1) came out—<continue>,

so that I did not need to distinguish between em and en dashes in the same immediate context (much like doing the same with en dashes any hyphens). This isn’t exactly an everyday occurrence, but it nonetheless may be worth covering. JeffConrad (talk) 01:58, 12 June 2011 (UTC)

Yes, I would also probably prefer to avoid the issue by choosing a different punctuation in the middle of a sentence. But that rather supports my point – that the use of the en dash in the film title is not just a regular case of the sentence interruption function that the em dash is described as performing, but is a different function which we don't currently list, and should (and which would possibly be best performed consistently by a spaced en dash rather than a range of alternatives).--Kotniski (talk) 15:48, 12 June 2011 (UTC)

In that case, I think the italics (and the blue link, if present) make the intent clear regardless of how the title is punctuated. ― A. di M.​_plé​^dréachtaí 16:24, 12 June 2011 (UTC)

I think I would describe this as a "graphical separator" or some such term, and consider it a generalization of the usage proposed in points 4a/b of the poll. Do any of the style guides have anything to say along such lines?--Kotniski (talk) 14:15, 14 June 2011 (UTC)

To try to sort this out, I've added a question to the poll (numbered 6.1). Answers welcome (preferably at the poll page).--Kotniski (talk) 08:02, 17 June 2011 (UTC)

Personally, I prefer a comma or parentheses, but either an em dash or a spaced en dash would work.

BTW, an em dash isn't just an informal equivalent of a colon, it's the opposite of a colon. With a colon, the following material explains or itemizes the preceding text; with a dash, the logical order is reversed. — kwami (talk) 10:32, 17 June 2011 (UTC)

Can you add your views to the poll page (question 6.1)? As to preferring other options (comma, parentheses), so do I, but this tends to concern things which (like the Harry Potter film) are written with a horizontal line as separator in the real world. I think there are also some political organizations in some countries that have names like this (will look for examples).--Kotniski (talk) 11:01, 17 June 2011 (UTC)

...Things like Indonesian Democratic Party – Struggle, though I'm not sure why that particular one is punctuated that way.--Kotniski (talk) 11:08, 17 June 2011 (UTC)

An em dash isn’t just an informal alternative to a colon, but if we are to believe S&W, CMOS, MWM, GMAU, MLA, OSM, and NHR, it does also take that role. A colon is frequently used to separate a work title and subtitle (e.g., CSE), and the usage above would seem to be essentially the same—I think the separator could be a comma, colon, em dash, or en dash (spaced or unspaced). I disagree that we should always space en dash in this context, because spacing is the exception rather than the rule. The “two dash″ rule (whether the dashes are spaced or unspaced) would probably proscribe the example we toyed with above. JeffConrad (talk) 01:33, 18 June 2011 (UTC)

"Two dash rule"? Can you explain? And which example are you referring to?--Kotniski (talk) 10:24, 18 June 2011 (UTC)

Many guides imply no more than two dashes in a sentence, and GMAU and NHR explicitly so state. CMOS16 goes a bit further, stating

To avoid confusion, the em dash should never be used within or immediately following another element set off by an em dash (or pair of em dashes). Use parentheses or commas instead.

This seems common sense; one alternative that I indicated above would be

<beginning of sentence>—when the film Harry Potter and the Deathly Hallows–Part 1 came out—<continue>,

though I think a comma, colon, or parentheses would be better because with some typefaces the em and en dashes can be hard to distinguish. With spaced en dashes used in place of em dashes, I think one of the other marks would be essential (or the sentence should be recast). JeffConrad (talk) 22:44, 18 June 2011 (UTC)

From poll page question 6.1

Dick, we see all manner of things, but I’m not sure use necessarily argues for approval; were we completely descriptive, the MOS would serve no purpose. I’ve always wondered about MWM’s curious use of the spaced hyphen in some addresses (e.g., 2018 - 14th Street); its predecessor MWSM called for a space hyphen or an (unspaced) en dash (e.g., 2018–14th Street), a situation we haven’t yet covered here. I never really cared for either approach, but can probably count the number of times it’s mattered to me on one hand. Spaced hyphens are common in online versions of newspapers, and in the US, they usually represent spaced em dashes. In some cases (e.g., SFGate), publishers apparently had problems with writers entering non-ASCII characters using encoding other than Unicode (typically, MacRoman), and got gibberish with many browsers, so they discouraged any non-ASCII characters in online content. Most browsers have supported Unicode for many years, so the “web safety” issue is overblown. Recognizing this, some newspapers (e.g., the Washington Post) have gone back to using normal typographical characters in the online content. But I digress.

I think we both recognize that some editors will not use dashes, though I’d be more inclined to say “don’t want to be bothered entering” than “don’t want to learn to enter ” (how hard is it to click a link below the edit window?). In an MOS that prefers typewriter quotes, it would seem silly to be hypersensitive to other non-ASCII characters. But it would be nice if the intent of a spaced hyphen were unambiguous; if it’s to indicate a spaced en dash, so be it, and we probably should say so. Ditto for two hyphens to represent an em dash; banning this, as we currently do, just encourages those who won’t enter dashes to use characters we might not immediately recognize as dashes. JeffConrad (talk) 21:38, 18 June 2011 (UTC)

We should of course keep in mind that if the usage isn’t clear, the choice of symbol doesn’t matter. JeffConrad (talk) 22:30, 18 June 2011 (UTC)

I think the idea of "banning" is not what was intended. If the MOS says it's not the right style for WP, that means editors are encouranged to correct it, not that editors are banned from entering it. I'm just trying to recognize a way that we can all get along, since there are those who don't want to be inconvenienced by understanding dashes. Dicklyon (talk) 03:29, 19 June 2011 (UTC)

Though I don’t like spaced hyphens, I’ve never suggested banning them, and I’ve previously indicated my disagreement with banning the double hyphen to represent an em dash. But if we want to accept two hyphens for an em dash, a spaced hyphen for an en-dash alternative to an em dash, and a hyphen (spaced or unspaced) for an unspaced en dash, we should say so—the more easily an editor’s intent can be recognize, the easier the process of copy editing. This, of course, would address only the editor who does not want to bother with non-ASCII characters yet has managed to consult the MOS to find the “typewriter” approximations (I think manuscript or ASCII might be more appropriate, especially for editors who’ve never seen—or perhaps even heard of—a typewriter). Alas, I think some of us date ourselves . . . Anyway, if memory serves me, the issue we′ve been discussing is when (and perhaps whether) it’s acceptable to replace hyphens with dashes in the course of copy editing. So I think we should resolve those issues, and perhaps also suggest ASCII alternatives to the characters preferred by the MOS. JeffConrad (talk) 06:26, 19 June 2011 (UTC)

Reminder concerning "how hard is it to click a link below the edit window?": It's about as hard as it is to click the closely similar link above the edit window, which you were unaware of until I pointed it out to you. My guess it that it takes on average about a thousand edits before an editor could make a dash if he wanted to. And he has no reason to want to, because it takes more like ten thousand edits before he discovers that there are those who prefer dashes in certain contexts. And when he learns, he probably learns by imitation, not by reading through all that "disjunction" stuff. Art LaPella (talk) 03:55, 19 June 2011 (UTC)

Yes, I think you're right. I'm sure I was more than a thousand edits in before I discovered there was an MOS. And I didn't discover those insert buttons until a while later. I still don't use any of them, since it's easier to imitate (copy and paste) than to search for a button that I don't know exists. On the other hand, I started out knowing how to type a dash, being a long-time Mac user, so that was never something I thought twice about. Dicklyon (talk) 05:41, 19 June 2011 (UTC)

(after edit conflict) But I was entering the typographical characters from day one nonetheless. Because I was used to the links at the bottom, I guess I had no reason to look at the top for a replication. I probably found the links at the bottom on the first edit for which I needed a Unicode character (when I noticed that most editors didn’t seem to like the HTML named references I was using); admittedly, I discovered them before the drop-down box was added, so the links were more obvious. Moreover, I’ve used many non-ASCII characters for years, so I probably had a reason to look for how to enter them here. Would I quickly find either of these sets of links if I were starting today? We’ll never know . . . We should note that we now have the articles WP:How to make dashes and Help:Entering special characters (which exists mainly thanks to your pushing for it), so the question of how to enter dashes or other special characters should be less of an issue. Perhaps the MOS should direct the reader to these articles. Of course, an editor may first need to find the MOS to learn that there is such a thing as an en dash, and where it should be used. I found it fairly quickly, but I perhaps just tend to look for such things. This may simply say that we need to do a better job of directing new editors to the MOS.

Incidentally, I don’t think it’s nearly as hard to use the links at the bottom, simply because the most common symbols are directly available. If I had frequent need for accented characters, my response might be different. Like Dick, I only use the links when an easier alternative isn’t available. For the common characters, I rely on a custom keyboard layout, and on Alt codes for a few others (in Word, I rely more on AutoCorrect). JeffConrad (talk) 06:26, 19 June 2011 (UTC)

I wonder if we (and I′m as guilty as anyone) aren’t getting carried away with issues that are at best peripheral to the main topic here. Tony said of the spaced hyphen, “It would be the last on my list of allowables”; I assume he meant simply that this should not be one of the MOS’s “preferred” styles, and that consequently, any editor would be justified in replacing it with something more suitable, and an editor who objected to such a replacement would normally be out of line (Am I at least close?).

Dick, Art, and I have used varying characterizations of those editors who do not enter actual dashes; Dick and I contrasted those who don’t want to learn how to enter them and those who simply don’t want to be bothered with doing so. Art noted the additional group of editors who don’t know how to do it and may not even know that there is anything to know. Clearly, there is little we can do for an editor who doesn’t think to look for an MOS other than to make whatever cleanup we have time for. But if an editor does make it to the MOS, notes that we want em dashes and similar characters, has never heard of such things, and does not get sufficient guidance on what they are, when to use them, and how to enter them, we have failed. It should not be that hard to fix this with a few suitable WLs. Though I stand by my statements on the difficulty of entering such characters, especially with the articles mentioned above, there still will be editors who, for whatever reasons, will not bother with characters not on a standard keyboard; for those editors, we’re probably better off guiding them to keyboard equivalents than leaving them to do whatever comes to mind.

So I think we should try to cover all three groups:

1. Those willing to learn uses and entry methods for what′s ultimately preferred. We can describe what’s preferred, and direct them to articles on entry methods.
2. Those willing to at least learn the uses, but possibly unwilling to bother with non-keyboard characters. We can give them the same information as the first group, and perhaps give some keyboard alternatives. Many style guides have statements like “if the en dash is not available on your keyboard”; these have always struck me as a bit silly, because almost no keyboards provide these characters (unless we count Alt codes and Option codes). Perhaps we could simply state possible alternatives as non-judgmentally as possible.
3. Those unaware that there even are certain uses. For those who somehow make it to the MOS, the least we can do is try to clearly describe the uses, and guide them to articles on entry methods, so that those editors can intelligently decide how far they wish to pursue it.

More succinctly, I think we should

1. State what we ultimately want (provided we can find reasonable agreement on what that is).
2. Direct editors to information on how this can be done.
3. Give some alternatives for those who find doing it too much trouble.
4. Recognize that for editors who never make it here, there is nothing we can do, and waste no more time worrying about it.

JeffConrad (talk) 09:41, 19 June 2011 (UTC)

Sounds sensible, but I don't really think number 3 ("Give some alternatives for those who find doing it too much trouble") is necessary or helpful. In particular, I don't think I've run into situations where people have entered things stranger than a hyphen or two where a dash is intended. More options here will just dilute the point that we have a preferred style. On the flip side, we don't want to say what some alternatives are and then "ban" them. There's just no need to talk about things other than the style we're aiming at. Dicklyon (talk) 15:04, 19 June 2011 (UTC)

As I say in more detail elsewhere, it′s easier to correct something if the thing to be corrected is used consistently. I guess I don’t see a problem with removing any mention of double hyphens, but I’d buy the claim of proper typography more if we applied it consistently. JeffConrad (talk) 02:10, 20 June 2011 (UTC)

"WP:How to make dashes ... Perhaps the MOS should direct the reader to these articles." It does. It's the second sentence of WP:DASH. Art LaPella (talk) 18:16, 19 June 2011 (UTC)

We cover the specific article, but not the more general one, which is arguably more useful overall (I somehow can’t see WP:How to make section signs and so on). Ostensibly, this discussion is about dashes, but in most of the immediately preceding comments, we could easily replace dash with almost any non-ASCII character. I hate to write the program that tracks exactly 17 salesmen. JeffConrad (talk) 02:10, 20 June 2011 (UTC)

And number 4 needs to be recognized so we can aim dash advice at copyeditors, especially automated copyeditors, not just authors. Art LaPella (talk) 18:19, 19 June 2011 (UTC)

I am like a person in the 3rd world who values physical security over political freedom

Personally, I find the differences tiny and the work to enter them a pain in the ass (Word is much easier to work with...Wiki is such a pain in the ass to actually compose content for). Telling me to type alt-gobbledigook is just not worth it.

And I think the amount of time we have tied up in this crap is soooo not worth it. how much marginal value have we gotten from dashes being differentiated in articles versus if Tony would teach people to structure their thoughts better in writing? I think any style guide will put way more emphasis on organizing content to have a strong structure than on dashes. And Wiki is just atrocious in terms of people organizing content. You can't gnome your way to an organized product, one that groups things into the best groupings, delivers points in the best order, eliminates redundancies, etc.

So...I really think Tony has done a disservice with this dash thing. That said...I think PMA has been even worse with the dead-ender guerilla warfare. Let's do it one way or another and just ROLL!

TCO (talk) 21:51, 10 June 2011 (UTC)

Oh...and I HATE the idea of practice varying from article to article. That is what makes referencing a mess. I don't care which system we use, but having all, REALLY means having none. "Consistent in article" is a crock of shit. Just check some. And causes huge inefficiencies for users as they go from one to the other. Believe it or not, there are people who want to go research things (http://dl.dropbox.com/u/349981/fluorine_yellow.pdf) and write articles (Fluorine). Dealing with the revert warriors on style is just a total buzzkill.

Oh...and I hate the idea of "follow the RS". We should follow the RS on facts. But not on grammar (unless an article about grammar or in deciding our own style...and then still we adjuticate amongst different options and pick our path). But we should not go ALLCAPS in some journal article reference because the article had that in it's header...just as we would not have to follow their font. We follow them on "is F yellow" but not on centre versus center or footnote style or some grammar usage. Plus if we did...that puts us into having a bazillion styles again! TCO (talk) 22:13, 10 June 2011 (UTC)

Careful, don't spend too much time here, or you'll start sounding like the rest of us! :) Art LaPella (talk) 22:42, 10 June 2011 (UTC)

I know, man, all this unimportant time-wasting stuff is so boring, waste of energy, gets in the way of doing more important editing work. Let's just all get along and get on with it. Oh, and I just HATE the way other people want to do it differently than I do; why can't we just ban all of them, and get back to what's important, which is doing it my way. Just don't treat any of this as important, and I promise I'll make the problem go away. Come on, people! Who's with me on this? Don't waste your time disagreeing or agreeing, but if you think what I said is irrelevant to anything you care about, pile on! And don't forget to say who, of all the people talking about things that don't matter, has done us the worst disservice. I want your vote! Dicklyon (talk) 01:21, 11 June 2011 (UTC)

Naw, you can't have my vote for the "worst disservice". You're perhaps the most extreme example of a prescriptivist faith I don't share, but you're civil about it, and you give every appearance of believing your own speeches. If you were the only guy I had to deal with, this place would run a lot smoother. Art LaPella (talk) 04:37, 11 June 2011 (UTC)

What then does Art propose for us all? This is no idle question; I'd really like to know. We can agree that there are far too many pages of MOS (I only take WP:MOS and WP:MOSNUM seriously myself; and I make damn sure I know the content of WP:MOS). We can agree that we need to trim things down drastically within WP:MOS itself. But you too would want to preempt disruptive exploitation of lacunae in black-letter guidelines, yes? And in the present voting we've seen how bizarrely people can interpret even the most "commonsense" summary wording. The prescriptivist–descriptivist dichotomy is laughably simplistic, and a disgrace within the populist linguistics industry. So, ah ... what's your bag, Klute? Noetica^Tea? 05:20, 11 June 2011 (UTC)

"far too many pages of MOS ...trim things down drastically within WP:MOS itself". Sure. We would probably get a lot more compliance with an MoS we can actually expect people to use, than what we would lose from the missing rules.

"preempt disruptive exploitation of lacunae in black-letter guidelines" To some extent yes, but that leads to chasing lacunae within lacunae within lacunae. I'll take your word for it that the "prescriptivist-descriptivist dichotomy" is something more complicated than what it sounds like.

Philosophically, I would first emphasize that the goal is to help Wikipedia spread knowledge, and not to demonstrate how many rules we know. That sounds uncontroversial until I apply that philosophy to specific proposals. Some of my less radical ideas, in addition to the above:

What I think is more important than the number of rules, is how easily people can find them. So in the absence of radically cutting subpages as described above, the search box in the upper right corner of the main MoS page is an underappreciated way to navigate the rules, especially rules on subpages. I would move and expand that box to become the main focus of the main MoS page. Everything else on the page distracts from it, so I would move almost everything else to other pages. Remember, everything on one page would be half a megabyte; it wouldn't fit.

Another corollary of focusing on the ultimate goal, is that the Manual of Style talk page should be more about things like AWB regex code than about dashes. No matter how vital we think dashes are, it doesn't matter unless we have a way to get those dashes to 6,823,161 articles. Art LaPella (talk) 03:55, 12 June 2011 (UTC)

I agree with Art about the difficulty in finding relevant elements of WP policy, and not necessarily just those related to the MOS. But to simplify the discussion, let’s stick stick with matters of style. Unlike with most other style guides, there is no sure way to find what controls. In other style guides, one can usually find what’s needed pretty quickly from either the TOC or the index, and guides available in electronic form additionally offer full-text searching that is restricted to that guide. For Wikipedia, it′s much more hit and miss, and finding the information (especially every relevant part) is sometimes a matter of making a lucky guess in the search. For en dash, it seems to work fairly well—the search finds the article, which directs an editor to that section of the MOS. For some other topics, especially when there is no relevant article—an editor may need to sort through all manner of extraneous material to find the answer. As Art and I have previously discussed, this may just be the nature of a wiki, but that recognition does nothing to lessen the difficulty of finding the answer.

Certainly, the overarching task is providing accurate, useful information, and we probably do sometimes lose the forest for the trees. So what do we do? I hardly think “Do whatever you want with no rules whatsoever” is the answer, because where would that stop? Just en dashes? Just punctuation? Just style in general? I suppose one approach would be to direct the editor to follow a major style guide like one of those in the list and apply it consistently in any article; we could further remind that MOS:STABILITY controls, so that any reasonable application of a major style guide should be left intact. But what of the editor who finds some obscure (and by this, I mean truly obscure) guide that deviates significantly from practice of most other guides. And what of the editor who doesn’t have ready access to any guides and doesn’t wish to buy one or make frequent trips to a library. As we’ve repeatedly said, we usually overlook almost any marginal usage that doesn’t amount to bad faith, with the expectation that some other editor will some day clean it up. Of course, it would be nice to have some guidance readily accessible to minimize the required cleanup. So perhaps we get back to making what we do have easier to find, though I have no immediate suggestions to offer.

As for focusing discussion here on REs that might automate copy editing, I think we’d quickly lose the vast majority of the participants. And relying on REs to fix much other than number ranges seems like a tall order, even with some of the more complex (and esoteric) enhancement to Perl REs. In particular, even if some magic bullet could be found, I fail to see how it could distinguish two-person endeavors from hyphenated last names. Some things are simple; as mentioned number ranges (including scores and votes) are usually easy to match, though the sweep may be so broad as to catch some instances of the minus sign by mistake. And many attempts at em dashes are similarly easy to recognize: two consecutive hyphens between groups of alphanumerics (or delimited by spaces)—I think this may be why we sometimes find real em dashes in online versions of newspaper articles in which almost everything is plain ASCII. With a single spaced hyphen, it’s a bit tougher, because we can’t tell whether it’s meant to be an em dash or a spaced en dash (or course, most US newspapers could simply require this to be an em dash. The TeX conventions of -, --, and --- for hyphen, en dash, and em dash would be easy to recognize if all editors knew (and agreed on) how to use them, and I think we know how likely that is, especially when so many editors seem unaware of even the two-hyphen convention for em dashes. So while bots and AWB can assist in cleanup, we need to recognize their limits, and probably accept that many articles may not follow the MOS to the extent that we would prefer. And I say this as one who has written several similar programs to add tagging to ordinary text, and which worked quite well in fairly limited circumstances (I only attempted en dashes between numbers). Of course, I may just not know what I’m doing . . .

Whether we rely on automation, the editor, or combination thereof, we still need to find what we can agree on for rules. And I don’t think “no rules” is the answer. JeffConrad (talk) 07:01, 12 June 2011 (UTC)

My AWB attempts to automate more (but less than half) of the MoS than anyone else I know of, although I have been concentrating on Simple English lately. And I haven't tried to use anywhere near all the tools the AWB manual describes; I'm just persistent. Of course I can't "distinguish two-person endeavors from hyphenated last names", but I can identify candidates for various kinds of edits. No two-person endeavors, but numeric ranges in ordinary text are just one of maybe 50 categories I identify, and I only expect to actually change around half of those candidates. Some Manual of Style editing is done manually, so the following is an exaggerated to make a point: if it can't be automated it can't be done on a significant fraction of Wikipedia, so if you can't follow the automation discussion you aren't helping much. Once again, that is exaggerated.

Whether "no roolz" is the answer would be an interesting philosophical debate, but even if it's the answer it obviously isn't the consensus. To some extent, consistent spelling and punctuation allows for faster reading, but I think the utility of half a megabyte of MoS rules is wildly exaggerated. The main objective is to avoid confrontations like this page. Art LaPella (talk) 19:23, 12 June 2011 (UTC)

"the search finds the article, which directs an editor to that section of the MOS" I didn't mean the search box in the extreme upper right corner. I meant the search box above the "Search the MOS" button at MOS:. Art LaPella (talk) 19:41, 12 June 2011 (UTC)

I never noticed . . . the things you learn every day (even though the button is pretty clearly labeled, it would make more sense if “Search” were “Search the MOS”).

We may just have differed on terminology; I never considered an interactive search and replace “automation”, but I guess it is to some extent. I’m not mainly a copy editor and have no intention of becoming one, but I probably also should learn AWB, if for no other reason than to understand discussions like this (and perhaps clean up my own stuff).

Half a meg? Not really that much anymore (I wonder how big CMOS is?). I do tend to agree that it should contain everything necessary but no more. Alas, not everyone has the same idea of what is necessary.

Yes, almost any use of time would be more productive than confrontations like this page. JeffConrad (talk) 04:30, 13 June 2011 (UTC)

I recomputed the half a meg I added last year. This time I got 1,423,995 bytes, not less. So we must be looking at different numbers. I used the {{Style}} template for a list of subpages. That list doesn't include WP:JEW/MOS which has the words "Manual of Style" in its name but not as a category, so it isn't on the Style template. If I should have counted it, there are probably other subpages in the same situation. I used the figure at the top of each edit history page as a byte count. My total:

Boring statistics

MOS: 153,685. MOS:AM 18,883. WP:CMOS 84,462. MOS:FILM 42,706. WP:NOVSTY 21,325. MOS:TV 43,948. WP:VAMOS 30,194. WP:WAF 35,484. MOS:MUSIC 29,064. WP:SAMPLE 4,736. MUSTARD not counted. MOS:CHARTS 7,322. WP:ACCESS 26,175. MOS:BIO 24,368. WP:LINK 38,273. WP:SELF 12,219. WP:WTA 14,250. WP:ABBR 15,395. MOS:CAPS 27,000. MOS:NUM 86,759. WP:PRON 29,253. MOS:PN 7,295. MOS:SPELL 34,392. MOS:TEXT 15,473. MOS:TITLE 7,674. WP:CAPTION 11,164. MOS:ICON 24,131. WP:EMBED 16,121. WP:WORKS 9,934. WP:LIST 26,520. MOS:RJL 30,701. WP:STAND 15,503. WP:FOOT 45,587. WP:LAYOUT 23,338. WP:LEAD 28,745. WP:TABLE 22,270. WP:TRIVIA 6,627. WP:MOSLAW 4,118. MOS:TM 5,924. MOS:CA 36,778. WP:FRMOS 16,302. MOS:HAWAII 10,055. WP:IMOS 13,625. MOS:JAPAN 33,313. WP:MOSSG 8,954. WP:MOSPOL 7,455. MOS:ISLAM 14,139. MOS:LDS 26,282. WP:MOSCHEM 17,206. WP:MOSCOMM 7,648. MOS:MATH 44,215. MOS:MED 49,778. WP:AMB 8,962. WP:MOSDAB 35,380. WP:IBX 12,885.

Total 1,423,995 bytes. If those subpages aren't the real Manual of Style, they shouldn't be allowed to use those words, and they shouldn't call themselves a WP:GUIDELINE. Art LaPella (talk) 06:30, 13 June 2011 (UTC)

Art, I have verified your addition just now. Useful information. But is your list up to date? See the list I have derived at WT:MOS, for the RFC on moving to subpages. You list 54 pages; I make it 58 (including WP:MOS itself). Noetica^Tea? 07:13, 13 June 2011 (UTC)

My list (which also includes WP:MOS) was derived from {{Style}}. That's all I know, in addition to the fact that being part of the Manual of Style is a semantic question, requiring a definition if you want precision. If the Style list is wrong, perhaps that template should be edited and/or (forbidden word) the extra four articles should be added to the category. Art LaPella (talk) 13:58, 13 June 2011 (UTC)

I don't think of myself as particularly prescriptivist; I wonder why I come off that way. I've repeatedly said that no editor is required to use en dash; I've suggested rewordings to get away from "requires". I've supported the idea that the "usages" of en dash described in the MOS say what an editor is "permitted" to do. I've agreed with the "stylistic alternative" for em dashes that some others have objected to as not prescriptive enough. Oh, well, I guess defending the MOS can make me look prescriptivist. To me, the prescriptivists are the ones saying we can't use en dashes, because many sources don't; or because they're hard to type; or because they can't see the difference; none of which seem like sensible reasons for them even to care, yet they care enough to object to us using en dashes; go figure. Dicklyon (talk) 06:53, 11 June 2011 (UTC)

Few people seem to be saying that nobody can use endashes; some people say that they themselves won't. But if you really want to get away from "requires", please do a draft; genuinely removing requires will do much of what I would like to see. Septentrionalis PMAnderson 14:40, 11 June 2011 (UTC)

The only "required" is in "the independent status of the linked elements requires an en dash instead of a hyphen," for which we seem to have good rewording proposed already. The rest is a description of the recognized roles of dashes, and some instructions on preferred use of spaces around them (spacing, and whether to make it more or less prescriptive, is where much of the disagreement is, but not one of my big issues). We can add that it is specifically OK for editors to use hyphen if they don't want to enter dashes (but not OK to undo the work of editors who upgrade to dashes, unless that is done in error or in a way not supported by the described uses). Dicklyon (talk) 16:36, 11 June 2011 (UTC)

not OK to undo the work of editors who upgrade to dashes, unless that is done in error or in a way not supported by the described uses. There is your prescriptivism in a nut-shell.Septentrionalis PMAnderson 16:45, 11 June 2011 (UTC)

(left) Changing a hyphen to a dash is often no upgrade, but a change from one variant to another ; in those cases in which a hyphen is overwhelming usage, changing it to a dash is deleterious. Do you stay with this position? Septentrionalis PMAnderson 16:45, 11 June 2011 (UTC)

I understand your position differs from most of us on that point. Most of think that it is OK for WP to have its own style, which says what variant is preferred, and to encourage editors to work toward that style, but without any requirements that contributions conform to that style. It seems to be a system that has been working well. Dicklyon (talk) 17:06, 11 June 2011 (UTC)

No, you do not understand my position; of the following four statements, you argue in favor of the one with which I agree.

1. Wikipedia should have a Manual of Style; I agree with this, as most commentators do.
2. Wikipedia's Manual of Style should have a preferred style. This is more dubious, but certainly true in many cases.
3. Wikipedia's Manual of Style should prefer what Dicklyon likes, because he thinks it's an upgrade. This is far more dubious.
4. Edits done to install the style Dicklyon likes should not be reversed. Not at all.

I will see what must be done about this. Septentrionalis PMAnderson 17:33, 11 June 2011 (UTC)

I agree with 1 and 2. Not the part about "dubious" in 2 though; I think having a preferred style is the whole point of the MOS. There's no reason to get involved with discussions of my own preferences. Earlier you said "please do a draft", but I'm going to avoid that, as you'd take it as just my own preferences; I'd like to see a draft that incorporates the constructive suggestions that came out in the votes above. Dicklyon (talk) 20:03, 11 June 2011 (UTC)

With regard to “overwhelming usage”, I am led to ask “According to whom?” If we are to believe Merriam-Webster, they have always been descriptivists in their dictionaries as well as in their style manuals, and reviewed 15 million entries in arriving at the the MWM. I’m sure only a small fraction of these were examples of en dashes, but I nonetheless wonder if anyone here has made a similarly comprehensive review of actual usage.

I guess I’d also like to know how using an en dash in a manner that finds reasonable support in major style guides is deleterious, other than that “PMAnderson doesn’t like it.″ If, as some here have suggested, the distinction is minor, or as Chicago have suggested, most readers will just see a hyhpen, what’s the problem? Those readers will see a hyphen whether it is actually a hyphen or an en dash, while those who notice and appreciate the difference will recognize the en dash. Thus post–Civil War years would read post-Civil War years for those who don’t see the dash; the reverse is not true for those who do recognize dashes. I don’t think anyone is suggesting that the sky’s the limit; I′d be among the first to object to post—Civil War years. But with either of the last two examples, we′d hardly chastise an editor who used them; rather (assuming we noticed), we’d simply change to the en dash. I simply cannot see the basis for getting bent out of shape over such an edit.

As for “New Dick’s Rules”, the votes suggest that either he’s running a lot of sock puppets or he’s hardly alone. We don’t seem to find complete agreement with the current MOS, but then we never find complete agreement on almost anything. I honestly was surprised by the strength of general agreement that we have found so far. JeffConrad (talk) 00:43, 12 June 2011 (UTC)

Jeff, it may surprise you to learn that you are one of my many sock puppets. But keep quiet about that. Dicklyon (talk) 00:56, 12 June 2011 (UTC)

TCO, I agree in part but probably disagree in greater part.

• Yes, the amount of time invested in this discussion boggles the mind, and the thought of having it run through 16 July goes even further. As I mentioned, it would seem that resolving the substance of the Arab–Israeli conflict pales in comparison with determining whether it should use a hyphen or an en dash.
• Just pick one way and roll? I’m all for it, as long as we pick my way. When the decision is made by consensus, the process is sometimes more difficult. Quite honestly, the main reason for the seemingly endless seems to be vehement disagreement with the current MOS by a fairly small group of editors (and I don’t mean “Dick Lyon and friends”).
• Inter-article variation? Yes, it would be preferable to avoid it. But if we sought to preclude it, we’d probably never be able to find consensus that would permit doing anything. And though many people seem to notice variations in punctuation and typography, those differences arguably pale in comparison with others such as English variations and writing style, which often exist even in professionally edited print publications.
• I agree that organization of ideas is a far greater determinant of article quality than whether we use a hyphen or a dash. But it’s an unavoidable consequence of any random multi-author effort (or is that random–multi-author effort?). Ultimately, it’s amenable to cleanup, but when such a revision is tantamount to a “gut and amend”, it is not done without controversy. And in any event, it’s an issue that we would encounter just as much if we didn’t invest another nanosecond on this topic.
• I think I agree on not primarily looking to “RSs” for guidance on style, because doing so would entail major research and endless argument for every article, possibly without resolution; the effort expended in this discussion would pale by comparison. The main reason for a style guide is to avoid repeatedly solving the same problem (often with different solutions).
• The effort to enter characters not on a standard keyboard? That’s of course a personal judgment. For me the effort is inconsequential; I′d swear that in the last discussion of quote characters, some editors expended more effort arguing why it cannot be done than I have spent in my entire life simply doing it, using Word as well as about half a dozen other applications. If I were to describe difficulties of various activities of Wiki editing, special characters wouldn’t even make the list. But even if an editor finds it too much trouble, he or she can simply enter a reasonable typewriter approximation and another editor will probably eventually correct it, just as many other things are similarly corrected. It has been said that a wiki is somehow different from conventional media because we don′t have copy editors. This isn′t really true; that function is simply fulfilled by certain other editors rather than by people specifically assigned to the task. And in any event, dashes are far easier to fix than drivel. JeffConrad (talk) 01:33, 12 June 2011 (UTC)

Jeff, we should be grateful that Dick's quest to undermine everything Wikipedia stands for has finally been exposed. I didn't dare speak up, because I was intimidated by his continued threats to "see what could be done about me", and by his crafty hijacking of my account to make it seem that I agreed with him. But seriously, I'm glad the focus is now back on TCO's well-motivated complaint.

Essentially I um, ... agree with you. TCO's frustration is understandable, but what is to be done? TCO is right to want consistency between articles. (This is fixed in policy at WP:TITLE, let's remember, as one of five principal criteria: "Consistency – titles which follow the same pattern as those of similar articles are generally preferred"). The call for consistency within an article does not mean related articles should go their own ways, with interminable replicated squabbling at each of them. TCO is right to call for consistent, practical, centralised standards. But you are right that this is not easy to achieve, and politics will inevitably intrude as people crusade – under the banner of their brand of Wikifreedom. Where I disagree with TCO, and with many commentators (including some admins from whom we expect greater insight and restraint) is how censure is to be meted out. Who is to blame, when debates rage on and on, over what outsiders see as trivial? Often a debate is offensive, or a disgrace to the Project, while no participant is at fault. Eruptions of disagreement in a participatory experiment on the scale of Wikipedia? Entirely predictable and unavoidable. This is not to say that no participant is ever blameworthy. It's just that we can't make the easy assumption that someone must be – and certainly not that everyone involved in the debate must be a fool or a troublemaker. Noetica^Tea? 02:57, 12 June 2011 (UTC)

I actually MEANT IT when I said I valued stability over "my side winning". I just want one side of the civil war to vanquish the other side so my kids can walk down the street...and I can operate my little teashop. I actually mildly disagree with TonyS's view on dashing and typography (I think the problems in edit mode and the amount of manual formatting is a real turnoff and the WMF agrees with me). But I value stability over "rightness". I will LEARN to alt gobledigook. Just want us all writing articles. And article to article difference is a charade. It's unworkable in practice for many reasons...and in practice leads to "within article" inconsistency. TCO (talk) 16:51, 12 June 2011 (UTC)

I guess given all we have going on, my preference is we go with the Tony style. That we just realize others will come after and clean stuff up--we really don't inhibit people from writing even if they double-hyphen their appositive dashes (like books from when I went to high school advise). All of the Tony side is pretty kind to people that just mess up, or are lazy, or the like. They just don't have time for the dead-enders hiding in spider holes. TCO (talk) 16:51, 12 June 2011 (UTC)

Let's be the bullet and have "approved style" as the typographically elegant one. We DON'T stop people who write drafts that are not perfect (really we don't). We just stop the style-revert dead-ender revert-warriors.TCO (talk) 16:59, 12 June 2011 (UTC)

Hopefully in time (along with other prompters, God bless the drive for ergonomics under the false flag of feminism) we get a user interface that makes all of this less painful (but let's have the pain first...to drive the change).

One minor thing is would appreciate it if the format lovers were churning out the content as well. They need to feel the pain when in edit mode and you can't read your words from all the footnote code being inline (I think wordprocessors have been able to do better than this for 20 years or so), etc. etc. (Yeah alt gobledi is no big deal...it's just there is a LOT of straw already on the camel).

In short, let's require a style which nobody outside Wikipedia actually uses, because a few people think it looks keen. Yeah, that makes sense. Septentrionalis PMAnderson 20:59, 12 June 2011 (UTC)

I hope TCO has found our new Wikipedia:How to make dashes, which helps the alt-gobbledy problem. Of course a problem still remains, mainly because people won't necessarily find that article when they need it. Art LaPella (talk) 21:52, 12 June 2011 (UTC)

I know the stuff is out there and can learn it. I would not minimize the PIA of manual formatting as the "straws add up" and distract from efficient consideration of content and organization. But this is really a separate issue and supposedly WMF is working on getting edit mode cleaned up. (It's shocking...I mean each desktop is a supercomputer now...this stuff CAN be automated. Look at Open Office even.) That said, I still think we should do it Tony's way. Even if I dork it up, someone will fix it later and no one is mean to me about not getting it right. PMA wants to play sealawyer and is not really unhappy about the work he needs to do. (He actually will change stuff back even after people fix it.)[That was User:TCO also. Please sign!–Noetica^Tea? 22:29, 12 June 2011 (UTC)]

Fine, I confess; I supported OLdSoeak, despite the clear intentions of Big Brother. ;-> Septentrionalis PMAnderson 01:07, 15 June 2011 (UTC)

See also the more general article Help:Entering special characters. I’m not sure there’s a really good way of entering special characters in any software I’ve ever used. A lot probably depends on what a person is used to, but in my experience WYSIWYGs are good for some things and less good for others. I agree that the edit window with any of the Wiki editing tools can be hard to read, and special characters are arguably a small part of the problem (and I started with markup languages long before I used a WYSIWYG). Of course, this isn’t really the main issue here—if we can’t agree on what should be entered, the difficulty of entry (or lack thereof) doesn’t matter. JeffConrad (talk) 04:14, 13 June 2011 (UTC)

• Please, no more talk of 11 December 1920–22 January 1921. There are millions of examples of the spaced dash between full dates on en.WP, and very very few that go against the long-standing rule. I'm willing to support a loosening of the rule aside from dates. Tony (talk) 07:56, 20 June 2011 (UTC)

Spacing of en dashes in ranges

This section is moved from Wikipedia talk:Manual of Style/dash drafting#Spacing of endashes. Ozob (talk) 10:47, 13 June 2011 (UTC)

• • Disagree wholeheartedly. I do agree that when one of the endpoints in a range contains spaces, then the en dash may be spaced (10 W – 100 kW, 5 January 1919 – 21 January 1919), but I don't think it's always necessary (25–30 mm, 4:30–5:00 pm, 3–6 November). But as far as disjunctive en dashes go, I see no reason to space anything. (Also I feel that I should point out a previous RfC on this topic: Wikipedia talk:Manual of Style/Archive 114#RfC: Disjunctive en dashes should be unspaced.) Ozob (talk) 12:41, 11 June 2011 (UTC)

    What? Not spacing the dash in 25–30 mm, 4:30–5:00 pm, 3–6 November is exactly what this is about. (In 25–30 mm the operands of the dash are 25 and 30, neither of which contains a space, mm being ‘factored out’. I agree that's not obvious from the wording, but the examples should make it clear. ― A. di M.​_plé​^dréachtaí 16:38, 11 June 2011 (UTC)

    I am not convinced that the English language factors out units. I interpret "25–30 mm" to mean "between twenty-five and thirty millimeters", and I interpret the latter to mean that there is an implicit, unspecified unit on the "twenty-five" which must be inferred from the context. I think the situation is clearer with "4:30–5:00 pm", which I interpret to mean "between 4:30 (am or pm unspecified) and 5:00 pm." 4:30 is likely to mean 4:30 pm, but it is conceivable that it could mean 4:30 am. For example, if we encountered "11:30–5:00 pm", then we would probably infer "11:30 am", not apply the distributive law of multiplication to deduce "11:30 pm".

    In any case, I generally oppose spaces. Ranges are the only exception I would like to permit: New York–Sydney, not New York – Sydney. I do not see any reason why our spacing rule for ranges must be the same as our spacing rule for other types of en dashes. The other uses of en dashes are completely different; why should they be spaced the same way? Ozob (talk) 20:59, 11 June 2011 (UTC)

    The units are understood to be the same. Do you have any coherent examples of the contrary? "25–30 mm" means the number of mm is 25–30, just as when we say twenty-five to thirty millimeters we mean the 25 is mm too. "4:30–5:00 pm" would have to mean 4:30 pm. You would not write "11:30–5:00 pm", you'd write "11:30–5:00" or "11:30 am – 5:00 pm". Unless the author was simply absent minded, which we should attempt to avoid. — kwami (talk) 02:21, 12 June 2011 (UTC)

    Most style guides seem to prefer the form 25–30 mm, but not everyone agrees, for the reasons Ozob states. Scientific organizations are quite fussy about clarity here; for example, NIST SP811 calls for 225 nm to 2400 nm or (225 to 2400) nm but not 225 to 2400 nm. Neither NIST nor the BIPM seem fond of range dashes, though offhand, I can′t find an example similar to this one in the latter’s SI brochure. I’ve generally followed SP811 in technical articles, but agree that they can look a bit ugly sometimes (e.g., 24 mm–105 mm lens, which isn’t commonly used by photographic equipment manufacturers, but there’s arguably some ambiguity in 24–105 mm lens). Beauty or clarity? Perhaps yet another issue . . . JeffConrad (talk) 03:04, 12 June 2011 (UTC)

    Yes, in technical lit, though I don't recall ever seeing parentheses like that in the body of a WP article. (Occasionally in a technical info box.) Assuming that will continue to be the case, there really is no ambiguity in "25–30 mm". If we use a spaced dash whenever either element has a space, then it would be absolute: if the mm only belongs to the 30, then the dash would need to be spaced. But in any case, no-one who's paying attention would ever use "25–30 mm" to mean 25 μm. — kwami (talk) 03:21, 12 June 2011 (UTC)

    (ec) Yes, exactly. If you were asked (say over email or IM), "There's a lecture 11:30–1:00pm. Are you going?", would you say, "Can I bring lunch?", "Why does it run all night?", or "I can't answer such a poorly phrased question!"? And the right answer to that question (there is a right answer) tells you that the English language does not factor out "am" and "pm". I argue by analogy that neither does it factor out other units. Therefore 25–30 mm is a combination of the independent elements 25 and 30 mm. Which (bringing us back around to spacing) tells us that there are times when the endpoints of a range contain spaces but the en dash should not be spaced. (I think we all agree that 25–30 mm is correct and 25 – 30 mm is not.) That's the point I was hoping to make, though I'm afraid it's getting lost. Ozob (talk) 03:31, 12 June 2011 (UTC)

    I tend to agree with kwami on 25–30 mm, especially about the form using parens, though in technical articles, I give considerable deference to recommendations from technical types. My position is more similar to Ozob’s with regard to 11:30–1:00 pm; though we probably can sort this out, it technically is incorrect, and in any event we should not need to backtrack to get it right. Just out of curiosity: are there publications similar to SP811 from the national standards laboratories of countries other than the US (preferably at the same price—free)? JeffConrad (talk) 03:44, 12 June 2011 (UTC)

    Well, “the right answer to that question” only tells us that human beings are good at correcting errors without even noticing them. Likewise, no-one would misunderstand me if I said “seeked” rather than “sought”, but that doesn't make seeked right, does it? (Now, I'd search language corpora to see whether usage such as 11:30–5:00 p.m. meaning ‘11:30 a.m. – 5:00 p.m’ is actually more common than can explained away as occasional slip-ups, but I don't know any corpus with an interface which can handle punctuation decently enough for that.) Also, you wouldn't actually use an en dash in IM, wouldjyou? :-) ― A. di M.​_plé​^dréachtaí 17:25, 12 June 2011 (UTC)

    I would like to agree to disagree, then, on whether this construction is erroneous or merely ambiguous. At least we are all agreed that it's bad! And I really would and do use en and em dashes in IM. But I'm on a Mac, where they're just option-dash and shift-option-dash. Ozob (talk) 21:59, 12 June 2011 (UTC)

    Actually, I just said why your arguments that it's not erroneous don't convince me, but I didn't actually argue that it is erroneous. I'm not convinced either way. ― A. di M.​_plé​^dréachtaí 16:05, 13 June 2011 (UTC)

    Whatever. My point was that I'd write 1400–3000 nm, 3000 nm – 1 mm, 1879–1955 and 14 March 1879 – 18 April 1955 this way; feel free to suggest a better way to express this, but I think that if the examples are included the intent is clear. ― A. di M.​_plé​^dréachtaí 16:23, 12 June 2011 (UTC)

    Exactly; and a "disagree wholeheartedly" on such a minor quibble ought to at least come with a proposed improvement; or be amended. Dicklyon (talk) 16:44, 12 June 2011 (UTC)

    Well, considering my past participation in wars over the whitespace next to short horizontal lines, I hope I can be forgiven for having an excess of feeling on this issue.

    I thought of something that I'd like to bring up here. What does everyone prefer here:

    30 May–5 September

    30 May – 5 September

    The reason I ask is because both endpoints have a space. In 30 May 2011–5 September 2011, there is (as we all know) the possibility that "2011–5" will be misinterpreted as a range. This is an argument for requiring spaces, as in 30 May 2011 – 5 September 2011. But that doesn't happen in the above example. So: Spaces or no spaces? Ozob (talk) 21:59, 12 June 2011 (UTC)

    I think all agreed with the rule that says spaces in that case. Dicklyon (talk) 22:22, 12 June 2011 (UTC)

    Then I would like to suggest the rule: En dashes used in ranges are spaced if both endpoints of the range contain spaces. Otherwise they are unspaced. Disjunctive en dashes not occurring in ranges are always unspaced. Ozob (talk) 22:42, 12 June 2011 (UTC)

    That might work; though I still wonder if there are other examples of ranges in which one side has a space and the other doesn't, beyond the examples where the "mm" or "PM" implicitly applies to both parts. If we start with the general condition that the en dash connects parallel elements, it's more clear that those can be factored out. In things like city names and person names we run into elements that sometimes spaced on one side or both; if we don't put spaces in New York–London, we also wouldn't put them in New York–Los Angeles, I would think. This suggests that the relevant test isn't whether both sides have spaces, but something else. Let's think more about the best way to pin down some pretty general cases, or perhaps give some advice about how to decide which is best without trying to pin down all the cases. Dicklyon (talk) 23:10, 12 June 2011 (UTC)

    It's not “both endpoints of the range contain spaces”: I would still use spaces in 387 – 17 March 493 (a range whose starting day is unknown but whose ending day isn't). Also, I'd write 50 – 100 million if the bottom of the range is 50 and 50–100 million if it's 50 million (even though I'd only ever use the former if the context makes that clear anyway, e.g. a list or table where the previous entry is 0.000025 – 50 and the next one is 100 million – 200 trillion). ― A. di M.​_plé​^dréachtaí 23:45, 12 June 2011 (UTC)

    That date range is a really, really good example.

    Also, it brings to mind 387 – March 493. I would space that, too.

    I don't know how to formulate a rule for these that does not also space 5–18 June. Ozob (talk) 00:21, 13 June 2011 (UTC)

    The relevant factor is whether we're linking like or unlike. In 5–18 June we're linking day with day, so there is no space. In 387 – March 493 we're linking year with month, so we do space. In 30 May – 5 September we're linking day-month with day-month, so we space both because we're linking spaced dates, and because the adjacent elements are unlike: month and day. In effect, 5–18 June is short for 5 June – 18 June.

    BTW, this is why "25–30 mm" can only mean 25 mm and why "11:30–5:00 pm" is simply wrong: for the latter, it can only mean the same pm, which means the same day, which would mean we're counting time backwards (from 11:30 pm to 5:00 pm of the same day), and no-one does that. I mean, if someone wrote (or said) "5:00–5:30 pm", you would never ask, "of the same day?", and if someone wrote "April 6–14", you would never ask, "of the same year?". And if someone wrote "April 14–6", you would assume it was a typo, or perhaps an odd convention for 6 o'clock on the 14th, but not an actual range of dates like *April 14 to May 6 or *April 14, 2010, to April 6, 2011. — kwami (talk) 13:36, 13 June 2011 (UTC)

    April 14–6 might be a seriously f***ed-up shorthand for ‘April 14–16’. Not that I'd ever use it myself, but I think I have seen something like that somewhere. ― A. di M.​_plé​^dréachtaí 16:05, 13 June 2011 (UTC)

    I think my position is correctly described by “En dashes in ranges are spaced if and only if at least one of the endpoints (not counting elements which are only written once but implicitly apply to both endpoints) contains a space.” But such language is strongly reminiscent of standards for programming languages. ― A. di M.​_plé​^dréachtaí 16:05, 13 June 2011 (UTC)

    Another example of a range where only one side contains a space: 2.5 ka – present (in Holocene). ― A. di M.​_plé​^dréachtaí 18:06, 14 June 2011 (UTC)

    So if I understand your position correctly, you are saying something like this: If two alike elements are being linked, and if both of them contain spaces, then the en dash should be spaced. If two unalike elements are being linked and either of them contains spaces, then the en dash should be spaced. Otherwise an en dash in a range should be unspaced. So you would write: 6–14 April (because you see this as two linked days), 30 May – 5 September (because both elements have spaces), and 387 – March 493 (because the elements are not alike). Am I understanding you correctly? Ozob (talk) 21:18, 14 June 2011 (UTC)

    I would not use a space in 30 May–5 September, and closed-up usage here is backed by every style guide I’ve ever read (though my familiarity is mainly with American guides and OUP). Probably what should guide is whether there are numbers on either side that could be mistaken for a range. I honestly don’t see how confusion could result from 5 September 2010–30 May 2011, though perhaps the initial impression could be disconcerting. This is a construction I haven’t found in any of my style guides, so I can’t say how they would handle it. With a construction that long, I would use 5 September 2010 through 30 May 2011. I find 387 – March 493 confusing with or without spaces; I would prefer some context (e.g., “from the middle of 387 through March 493” so that my initial reaction wasn’t “Huh?” JeffConrad (talk) 23:20, 13 June 2011 (UTC)

    It is somewhat realistic, though. In a medieval biography, you will often get a very complicated date range right at the start of the article. For example, Jerome ("c. 347 – 30 September 420"), Dante Alighieri ("May/June c.1265 – September 14, 1321"), Peter Damian ("c. 1007 – February 21/22, 1072"), or Albertus Magnus ("1193/1206 – November 15, 1280"). These constructions need spaces, as their endpoints are complicated things containing spaces. 387 – March 493 is on the edge of what needs spaces, but I think it ought to be spaced, too. Ozob (talk) 01:25, 14 June 2011 (UTC)

    The “c.” provides the context; spacing is probably reasonable. JeffConrad (talk) 02:58, 14 June 2011 (UTC)

[Outdent] I call on the participants (Ozob? A di M?) in this very productive excursus to move all of it to the discussion page, [...] Thanks for moving it, Ozob. Noetica^Tea? 21:25, 13 June 2011 (UTC)

I think we're trying too hard here to get a simple hard-and-fast rule. Sometimes leaving some discretionary wiggle room is a good idea. We could say that the en dash is usually unspaced, but can be spaced when connecting more complex items that contain spaces, as in June 2010 – April 2011 or 200 AD – present; it usually remains unspaced when connecting names, even if they contain spaces, as in Siefert–van Kampert and New York–Los Angeles. If there are other clarifying examples on which we agree, we can put them as examples. Dicklyon (talk) 21:30, 14 June 2011 (UTC)

I agree that covering every case would be difficult, but I think spacing is very much the exception rather than the rule, so I would focus on describing the exceptions. For example, I don’t see the need for a space in “June 2010–April 2011”, and all of my style guides seem to agree. The case for spacing “13 June 2010–1 April 2011” is probably stronger, but some guides nonetheless show this closed up. I don’t find this confusing, but again would probably prefer “13 June 2010 through 1 April 2011”. JeffConrad (talk) 23:56, 14 June 2011 (UTC)

OK, then for the date example, let's use the full dates as we use them in bios (I think everyone is pretty much in favor of the way we do those now); leave some discretion for other spacing exceptions. Dicklyon (talk) 00:14, 15 June 2011 (UTC)

A draft

Seeing that we seem to have overwhelming agreement on most of the issues, I've tried redrafting the whole dashes section based on what seem to be the likely results of the poll (although there might still be a change of decision on some points). My effort is at WP:Manual of Style/Dash draft - please tweak and/or leave comments.--Kotniski (talk) 14:26, 22 June 2011 (UTC)

K, thanks for this great start; I think it reflects most of the comments, and addresses most of the disagreements. I do have some suggestions: First, don't mix up the "dash" stuff in points 1 and 6 in the same list as the "long hyphen" uses of the en dash; use a separate subsection about that. Second, there are still two many distinct en dash uses listed. Merge points 3 and 5, since the numbers in 3 are just a special case of the situation described in 5. And on the ranges, point 2, include some non-numeric examples, like Monday–Friday. In general, I think we need more examples, to cover the range of things that we have discussed (like connected adjectives). Thanks again for a great start. Dicklyon (talk) 14:08, 24 June 2011 (UTC)

I tried to avoid subsections, for simplicity, but you're right that point 6 (once we get the wording sorted out) would be better placed after point 1. I'm not sure 3 is always a special case of 5, though, since in sentences like "they voted 5–4" we are not dealing with a compound modifier before a noun. An example like Monday–Friday would be good; I guess there are also examples like "they travelled London–Tokyo–Sydney" which aren't properly covered by any of the current points.--Kotniski (talk) 15:48, 24 June 2011 (UTC)

Ignores the fact that many editors (about a third of those polled) disagree with points 4 and 5; several editors would strengthen the recommendation to avoid compounded compounds.

Point 1 is badly phrased: the parenthetical use of dashes is a special case of the abrupt change – where it changes back. Septentrionalis PMAnderson 01:54, 13 July 2011 (UTC)

Categories

Oh, I've just remembered that we sometimes have disagreements about whether dashes should be used in category names (some people think there are technical reasons for sticking to hyphens in that case). Has this question ever been settled? If not, can it be incorporated into the current poll?--Kotniski (talk) 14:33, 22 June 2011 (UTC)

Is there a technical problem? Many categories do use en dash already. The only category discussion I recall was not about technical issues, iirc, but wanting to make the cat name match a recently un-dashed article name. I wouldn't say anything about it unless there's an actual issue. Dicklyon (talk) 19:54, 22 June 2011 (UTC)

And I just noticed that I missed your section above on "A draft", seeing only the latest diff. I'll take a look, but probably not right away, as I have a short trip tomorrow... Thanks for getting this started. Dicklyon (talk) 03:29, 23 June 2011 (UTC)

The "technical problem" (which I don't see as a significant problem, though I know from past discussions that some people do) is that "category redirects don't work properly" (so having a category name with a hyphen redirecting to one with a dash is not entirely satisfactory). To me the resulting problematic scenarios that people describe don't seem particularly problematic (compared with the problems that arise from having a mismatch between article adn category names), but it would be good to get consensus on this point so as not to have to keep repeating the argument whenever such an issue arises.--Kotniski (talk) 09:44, 23 June 2011 (UTC)

Should we clarify when not to use dashes for year ranges?

If we're redrafting our dash guidelines, there's a point I haven't seen discussed yet that could make editor's lives less contentious.

Currently, WP:YEAR states "Year ranges, like all ranges, are normally separated by an en dash, not a hyphen or slash: 2005–06 is a two-year range, whereas 2005/06 is a period of twelve months or less, such as a sports season or a financial year." When I edit sports articles, if I take this advice to heart, I'm usually reverted, or the subject of much haranguing on the talk page; it seems most sports editors are attached to the en dash for sports seasons. It might help if the section on the use of the en dash for year ranges echoed WP:YEAR. Perhaps it could even be made more clearly: That 2005–06 indicates the period from January 1, 2005 to December 31, 2006, and that 2005/06 should be used for a date range of less than that. CMOS16 ¶6.105 endorses, but does not require, this usage of a slash instead of an en dash.

Or, alternatively, we should clean up WP:YEAR to avoid dash-related confusion. Me, I like the rule. // ⌘macwhiz (talk) 02:25, 29 June 2011 (UTC)

Proposed new draft: plenary discussion at WT:MOS

Thanks are due to Kotniski and the others who have worked on a preliminary draft. I have opened a new section at WT:MOS for continued development of new dash guidelines, as 16 July approaches and we need plenary discussion with fuller participation. See also a summary of the action up till now, in the section that precedes that one ("Dash guidelines: toward a conclusion"). Noetica^Tea? 03:21, 13 July 2011 (UTC)