Wikipedia talk:Manual of Style/Dash draft

Hyphens in a list
"6. Spaced en dashes are sometimes used as separators, ..." Did you mean to say "but not hyphens"? Art LaPella (talk) 20:56, 22 June 2011 (UTC)
 * Yes, I guess that's the intention; maybe it should be said explicitly.--Kotniski (talk) 09:45, 23 June 2011 (UTC)

General comments

 * I thought there was reasonable support for eliminating mention of “typewriter approximations”. If the caveat remains, we should use the present tense, and possibly reword to something more contemporary like these are approximations used for ASCII text. If it’s mentioned, there’s a potential confusion between the two-hyphen typewriter convention for em dashes and the TeX coding of two hyphens for en dashes.
 * I've no objection to removing that whole sentence (I certainly wouldn't want to make it any longer).--Kotniski (talk) 09:59, 23 June 2011 (UTC)
 * Should the ndash template be mentioned for spaced en dashes? It handles the nonbreaking space with less effort and visual clutter.
 * Could be, though I was assuming that methods of entry would be dealt with at WP:How to make dashes, which is linked to.
 * (1) Would it not make more sense to use the more common parentheses rather than round brackets, especially since parenthetical is already used.
 * Fine by me (though in Britain, "brackets" is probably more common).
 * I usually say "round parentheses", "square brackets", and "curly braces" so that I can't be misunderstood on either side of the Atlantic. :-) ― A. di M.​plé​dréachtaí 20:46, 23 June 2011 (UTC)
 * Dunno about “round parentheses”, but I agree on the latter two. Whatever is said should be clear in BrE and AmE. JeffConrad (talk) 22:56, 23 June 2011 (UTC)
 * Changed to "parentheses". (The wording I used was taken directly from the present MOS.)--Kotniski (talk) 11:01, 24 June 2011 (UTC)
 * OSM claims they’re “parentheses”, but recognizes that they’re usually called “round brackets” in BrE. I suppose the way to please everyone would be to say “parentheses (round brackets)” or “round brackets (parentheses)”. I’d probably go with the former given OSM’s treatment, but either would be OK. JeffConrad (talk) 22:13, 24 June 2011 (UTC)
 * (2) I think we should allow December 1989–March 1990 as well as December 1989 – March 1990 because there’s little chance for confusion, and every guide I’ve ever read gives it the first way. That doesn’t mean it’s the only way, but to ban a reasonable common practice seems capricious.
 * OK by me, if that's what comes out of the discussion (this draft isn't intended to pre-empt any decision).
 * This could be implemented by changing “is spaced” to “may be spaced”. ― A. di M.​plé​dréachtaí 20:46, 23 June 2011 (UTC)
 * Seems OK to me. JeffConrad (talk) 22:56, 23 June 2011 (UTC)
 * So what do you think the consensus is – something like "always space the dash with full dates; optionally space the dash between other items containing spaces"? What about the 100 W – 10 kW example: is the spacing to be optional there too? --Kotniski (talk) 10:52, 24 June 2011 (UTC)
 * Seems to me closing up with full dates has the least support, though I don’t know if it makes sense to disallow US practice. For December 1989–March 1990, support seems split, so perhaps it should be allowed either way. As for 100 W – 10 kW, APA, CMOS16, and MWM would appear to close this up, though they don’t provide directly apposite examples. SP811 would appear to space this, though they deprecate range dashes altogether, preferring to or through. Is there an equivalent of SP811 that recommends implementation of the ISO 31 (or perhaps the ISO 80000) standards for the UK, Canada, or Australia? Perhaps it could shed some additional light. JeffConrad (talk) 22:13, 24 June 2011 (UTC)
 * FWiW, the EU style guide (sect. 3.15) agrees with what's being proposed. ― A. di M.​plé​dréachtaí 16:06, 25 June 2011 (UTC)
 * (4) I think it would help to suggest former prime minister as an alternative to ex–prime minister, which would usually not be used as an adjective, and would seem to have less justification for the en dash.
 * OK by me. Done.--Kotniski (talk) 10:55, 24 June 2011 (UTC)
 * (5) The senses should be in italics rather than quotes for consistency with (3), and so that it’s clear that they are words used as words rather than the sense of so-called.
 * Agree; done.
 * Well, it is usual to use italics for the words themselves and quotation marks for their meanings, e.g. ain’t is a well-known non-standard form (here meaning “haven’t”); here, I think it's the latter intention. Speaking of italics, do we need them in parenthetically, sharp break, independent and equal? I'd be happy to lose 'em. ― A. di M.​plé​dréachtaí 12:22, 24 June 2011 (UTC)
 * The call between italics and quotes is always a tough one, and I’m often almost tossing a coin—I guess either would work here, but we should be consistent. I’d shed no tears over the departure of the words in your penultimate sentence. JeffConrad (talk) 22:13, 24 June 2011 (UTC)
 * Of course, I meant I'd replace italics with quotes both in (3) and in (5). I can live with using italics in both, though. ― A. di M.​plé​dréachtaí 23:32, 24 June 2011 (UTC)
 * These are small details, but perhaps I had it right to start with with these quotes and italics – in the case of ranges, we really do mean the word to or thru (the dash stands for that word), whereas in the case of compounds, the dash only carries the meaning of the listed words. With scores, it could be interpreted either way. As to the italics on independent and equal, I'd say it was justified emphasis, though I don't insist. For parenthetically and sharp break, I used italics so that when we come to refer to the "two uses" in the next point, it will be a bit clearer what we mean (I tried to keep the layout simple, with a single list of bullet points for each numbered use).--Kotniski (talk) 10:28, 26 June 2011 (UTC)
 * (6) It’s not clear to me why the en dash is always spaced in something like Harry Potter and the Deathly Hallows – Part 1; this seems like a somewhat arbitrary choice. In a title, would we not normally do a the publisher did (e.g., Henry VI, Part Two)? I think a colon, comma, unspaced en dash, parentheses, or even perhaps an unspaced em dash could work just as well; I think the application to track listings would have almost as many reasonable options. I also think that without a link to an example to the use for track listings, most readers will have no idea what we’re talking about. Though it should be obvious, perhaps it should be mentioned usage within an article should be consistent. I don′t see a problem using different symbols for connecting track numbers to track times and for connecting musicians to instruments, but obviously usage for each should be consistent.
 * (6) “[B]ut no hyphens”. Would this not be treated the same way as “typewriter approximations”, however it is decided to handle them. It would seem to me that if we intend to deprecated spaced hyphens we should do so under WP:HYPHEN, and if absolutely necessary, at the beginning of WP:DASH. JeffConrad (talk) 08:03, 23 June 2011 (UTC)
 * I think the main reason for this point is to emphasize that we don't use hyphens here (i.e. that if we want a horizontal line, then it ought to be a dash). I don't see a problem with recommending spaced en dashes specifically, since they seem to be the ones that are already most commonly used in these positions, and recommending a single style would encourage consistency and discourage lame arguments.--Kotniski (talk) 09:59, 23 June 2011 (UTC)
 * Agree on the spaced hyphen, though I think it would make sense to mention it once or twice rather than several times (spaced hyphens are common in online newspaper content that uses only ASCII characters. In the US, it usually indicates a spaced em dash; in the UK, it may indicate a spaced en dash. It′s hard to say what it represents in a Google search . ..
 * Again, I think deferring to the publisher would be the right approach when possible (spaced hyphen possibly excepted); I just could not see writing Henry VI – Part Two because this would be uncommon usage. And for Kind of Blue, we could not use Miles Davis, trumpet and leader, as was done on the album cover. Even for Harry Potter and the Deathly Hallows – Part 1, I’m not sure the en dash is the most common—a Google search shows colons, commas, parentheses, and the Warner Brothers site has both a spaced dash and no punctuation (I’m not sure the spaced hyphen there is anything other a clueless keyboarder, and certainly would not recommend it here). Ostensibly, a single style might be nice, but it could be difficult to agree on what it should be. In particular, the spaced en dash seems a minority usage, especially in the US, where en dashes are nearly always closed up. JeffConrad (talk) 22:56, 23 June 2011 (UTC)
 * WP:HYPHEN already says "A hyphen is never followed or preceded by a space" (followed by irrelevant exceptions), and hyphens I find in Wikipedia's lists are usually spaced. But if someone wanted to object to me changing unspaced hyphens in a list to dashes, I don't know what guideline I would point to after this rewrite. WP:HYPHEN also says "A hyphen is used only to mark conjunction, not disjunction", but that's more of a koan than a guideline; one definition of "conjunction" is "a combination of two or more things", and that sounds a lot like what happens in a list. Art LaPella (talk) 00:41, 24 June 2011 (UTC)
 * By Dick Lyon’s suggestion, if the unspaced hyphen were used where the MOS called for an en dash, no further justification should be needed. If for some reason that doesn’t seem to grant sufficient authority, I guess we’d need to stick with the ban. Many other guides include “do nots”, so we probably could go either way. If we decide we need the explicit ban here, perhaps we need it for the double-hyphen convention, though as I’ve said, I’m happy when someone indicates an em dash with two hyphens rather than forcing me to figure it out. As for “conjunction, not disjunction” . .. JeffConrad (talk) 03:25, 24 June 2011 (UTC)
 * As I remember DickLyon's suggestion, editors can go ahead and use a hyphen, but they shouldn't complain when a copyeditor changes it to a dash and cites an appropriate guideline. That's the most we can expect anyway, no matter what we say here, and we might make that more explicit at WP:MOS because it applies to the entire MoS, not just hyphens. But it won't happen if we no longer have the appropriate guideline, in this case for an unspaced hyphen in a list. The proposed rewrite already has an explicit ban for double-hyphen: "Do not use substitutes for em or en dashes, such as the combination of two hyphens (--)", but not for single hyphens in a list. Not that I would really object to the hyphens, but the main goal is to get everyone into one big consensus. Art LaPella (talk) 04:08, 24 June 2011 (UTC)
 * There once was an essay saying something like, “If you can't write something perfect, write something reasonable rather than nothing at all; it can be improved afterwards”. This should definitely be mentioned in WP:MOS#General principles. ― A. di M.​plé​dréachtaí 12:47, 24 June 2011 (UTC)
 * Dick’s had quite a few suggestions, including “There’s just no need to talk about things other than the style we're aiming at”. The proposed rewrite retains the existing ban on two hyphens, which Dick’s suggestion would presumably eliminate. Perhaps I didn’t quite follow your earlier comment; were you concerned about an editor doing something like Harry Potter and the Deathly Hallows-Part 1? I don’t think that works. This example is now looking trickier than I had thought; for the title/subtitle, I think any of
 * Harry Potter and the Deathly Hallows, Part 1
 * Harry Potter and the Deathly Hallows: Part 1
 * Harry Potter and the Deathly Hallows (Part 1)
 * Harry Potter and the Deathly Hallows–Part 1
 * Harry Potter and the Deathly Hallows – Part 1
 * Harry Potter and the Deathly Hallows—Part 1
 * and perhaps even
 * Harry Potter and the Deathly Hallows Part 1
 * would be OK, but a hyphen, spaced or unspaced, doesn′t cut it. For a track listing, I think either of
 * So What? (9:25)
 * So What? – 9:25
 * would be fine; I’m much less sure of
 * So What?: 9:25
 * So What?–9:25
 * So What?—9:25
 * Perhaps I’m now agreeing with Kotniski that at least the track listings (and similar) may be a unique application. For the linking of musicians and instruments, I think it’s more like title/subtitle; any of
 * Miles Davis, trumpet and leader
 * Miles Davis: trumpet and leader
 * Miles Davis – trumpet and leader
 * Miles Davis—trumpet and leader
 * and maybe
 * Miles Davis (trumpet and leader)
 * would be OK. For a list like June 23, either of
 * 1990 – Moldova adopts the Declaration of Sovereignty of the Soviet Socialist Republic Moldova
 * 1990: Moldova adopts the Declaration of Sovereignty of the Soviet Socialist Republic Moldova
 * would seem OK, though obviously for a specific category like Day of the year we’d stick with a single established style.
 * As for “bans”: I think we should be consistent to avoid confusion and argument. If we make one ban explicit, we should do likewise for all. JeffConrad (talk) 06:28, 24 June 2011 (UTC)
 * Opportunities never cease . .. Conversion of units currently contains “Radiation - source activity”, of which I assume we would disapprove. To me, this would seem a candidate for Radiation—source activity or Radiation – source activity, but is it a “sharp break in the flow” of a title or does it “set off an amplifying or explanatory element”?.
 * On use in titles: if we include one alternative, we probably should include them all. But another approach would be similar to the proposed wording “spaced en dashes may be used between” [emphasis added], with the hope that editors would use common sense when using alternatives (Do I feel lucky?), and as long as there is no implication that the relationship between musicians and roles must follow the same format that between track numbers and durations. Would I have thought of this implication absent the current wording? I can’t say. JeffConrad (talk) 08:31, 24 June 2011 (UTC)
 * I think we're perhaps making too much of this rather marginal set of cases; we want to allow dashes and disallow hyphens in such cases, but not to disallow other punctuations. And encourage consistency above all. More examples of the kind of thing can be found in Category:Political parties in the Philippines – notice here we have several parties with horizontal lines used as separators (hyphens, spaced en dashes, unspaced en dashes) – surely we would want to recommend a single choice so that they all become at least consistent?--Kotniski (talk) 11:34, 24 June 2011 (UTC)
 * I had similar thoughts; more than anything, I think we wish to disallow spaced hyphens. But do we allow both em and en dashes (it would seem to me that we would for title qualifiers, which would make them not special usage)? We’d need to ensure that whatever we wrote would not likely be taken as precluding other punctuation. For your example above, it seems to me that it’s just a dash, unspaced em or spaced en; I don’t think an unspaced en works here (and I’m having doubts about it in Harry Potter, as well.
 * I do think we should add “set off an amplifying or explanatory element” or similar to (1), which would then cover Radiation—source activity; if necessary, we could specifically mention work titles as well, though again we’d want to make clear that we don’t preclude the more common colon or comma. JeffConrad (talk) 22:41, 24 June 2011 (UTC)
 * "we wish to disallow spaced hyphens" Once again, the "Spacing" paragraph at the end of WP:HYPHEN already disallows spaced hyphens. Unspaced hyphens are a problem only in dash draft point 6 because it says "optionally". Art LaPella (talk) 23:48, 24 June 2011 (UTC)
 * Why is (6) a problem? If spaced hyphens are never allowed, I don’t see how “en dashes may optionally be used” overrides; that en dashes may be used should not mean that spaced hyphens are an alternative. Are we all just too accustomed to spaced hyphens in lieu of what’s proper? I would delete optionally, however, because it’s inherent in may. JeffConrad (talk) 00:47, 25 June 2011 (UTC)
 * "en dashes may [with or without 'optionally'] be used" doesn't override WP:HYPHEN's prohibition of spaced hyphens. My point is unspaced hyphens in point 6. Unlike 1–5, which specify dashes, 6 allows for the possibility of other characters, as intended – but what wasn't intended is that those other characters could include hyphens. Most hyphens in situation 6 are spaced and thus outlawed by WP:HYPHEN. But if I find unspaced hyphens in a list, I would no longer have a guideline I could point to, if someone insisted that hyphens are the correct choice. I could say Jeff Conrad says not to use hyphens, spaced or unspaced, but the editors I encounter seldom care about what anybody else thinks. I also heard Kotniski say we're making too much of it, but at least get my point straight; it's about unspaced hyphens, not spaced hyphens. Art LaPella (talk) 04:41, 25 June 2011 (UTC)
 * “Jeff Conrad says not to use hyphens” would certainly strike fear in any miscreant’s heart . .. OK, so it is unspaced hyphens, which is why I asked above—guess it got lost in the other clutter. Nothing some editors might do would surprise me, but using a hyphen in such a situation (which is not one we mention in WP:HYPHEN) would seem mighty hard to defend. Have you ever had an editor object to replacing such a hyphen with a dash? We don’t prohibit asterisks or minus signs, either. I’m sure they’re much less likely than a hyphen, but where do we stop? I really don’t have that strong an opinion, though again, if we explicitly ban one thing, we really need to ban everything we intend to ban. I realize this is getting “legalistic”, but ultimately, such a reading is necessary if we are to avoid Humpty-Dumpty. JeffConrad (talk) 07:05, 25 June 2011 (UTC)
 * You've got me there; the only hyphen debates I remember are with Hyphen Luddites, who object to dashes in general. My most memorable AWB debates were about WP:NBSP, which would create the most radical change to our articles if we took that guideline more seriously. Art LaPella (talk) 18:38, 26 June 2011 (UTC)
 * As for “deferring to the publisher”, I think most book covers show subtitles by just using a line break and a smaller font (e.g. Pulp ¶ A Novel or ‘WHAT DO YOU CARE ¶ WHAT OTHER PEOPLE ¶ THINK?’ ¶ FURTHER ADVENTURES ¶ OF A CURIOUS CHARACTER, ¶ denoting a line break) Which of course is not viable in running prose. ― A. di M.​<i lang="ga" xml:lang="ga">plé​dréachtaí</i> 12:31, 24 June 2011 (UTC)
 * That’s why I included when possible . .. in fact, Warner Brothers seem to avoid punctuation on the cover of Harry Potter and the Deathly Hallows by using multiple line breaks. JeffConrad (talk) 22:41, 24 June 2011 (UTC)
 * My point was essentially that when possible effectively means ‘hardly ever’ in this situation. ― A. di M.​<i lang="ga" xml:lang="ga">plé​dréachtaí</i> 16:15, 25 June 2011 (UTC)
 * Well, I haven’t done a statistically significant survey, so I can’t say. The shorter the qualifier, the more likely the publisher is to avoid a line break, as in Henry VI, Part Two. Of course, publishers vary; this form seems to be the most common form in the US, but a quick search also turns up Henry VI: Part Two, Henry VI – Part Two, Henry VI Part Two, all sometimes with King prepended, and even the original title The Second Part of King Henry VI. A search also finds instances of spaced hyphens, but I assume they are just the result of CKs.* So perhaps “follow the publisher” invites the question “Which edition of which publisher?”. The variety of treatments here suggests that almost any of the forms I showed above are reasonable, except perhaps the spaced hyphen and probably the unspaced en dash (which I’ve just struck). I personally would prefer the simpler and more common colon or comma, but I think we should should allow any of the other reasonable uses. JeffConrad (talk) 22:59, 25 June 2011 (UTC)
 * &emsp; *Clueless Keyboarders.

A few more comments: ― A. di M.​<i lang="ga" xml:lang="ga">plé​dréachtaí</i> 21:03, 23 June 2011 (UTC)
 * I'd prefer minus sign rather than subtraction sign (but I can live with either).
 * Changed.--Kotniski (talk) 11:05, 24 June 2011 (UTC)
 * In point 4, I'd use different examples than the New York–Sydney flight etc. as these are covered by point 5 anyway. Also, I'd explicitly mention hyphen-based alternatives (pre-World-War-II technologies, ex-prime minister), which are more common in BrE. Also, for the suggestion to recast the phrase I'd use examples involving more obscure compounds, with real danger of confusion (pre–World War II technologies being nearly impossible to misunderstand as World War II is an extremely familiar ‘fixed’ phrase).
 * I'd swap points 4 and 5, as point 5 is arguably more similar to point 3 than point 4 is.
 * I can't see the point of mentioning image filenames in the MOS at all: they aren't normally written in normal English anyway (in today's FA there are File:Holy Thorn Reliquarywindow.jpg, File:Holy Thorn Reliquary 2407598204 8dda9afd97 o.jpg, File:Britishmuseumdunstableswanjewelsidecroppedclose.jpg among others), they are not normally ever displayed to non-editing readers, and most of the files are on Commons where the en.wiki MOS doesn't apply anyway. ― A. di M.​<i lang="ga" xml:lang="ga">plé​dréachtaí</i> 21:32, 23 June 2011 (UTC)

And a few more from me: JeffConrad (talk) 22:56, 23 June 2011 (UTC)
 * I’d strongly prefer minus sign, because that’s common name as well as the name for U+2212 (it indicates the subtraction operation). Alternatively, we could say something like “might be misconstrued as indicating subtraction”, but I think shorter is probably better.
 * I don’t think the New York–Sydney flight is quite the same as (5), because we wouldn’t use Chicago-Denver flight.
 * Huh? Unless you meant to use a dash in the latter example or to say “would” rather than “wouldn't”, I can't make sense of that. ― A. di M.​<i lang="ga" xml:lang="ga">plé​dréachtaí</i> 00:04, 24 June 2011 (UTC)
 * I hope we’d use Chicago–Denver flight to distinguish from an ordinary compound modifier, much as CMOS16 gives London–Paris train and NHR gives Dover–Calais crossing. JeffConrad (talk) 00:38, 24 June 2011 (UTC)
 * Yes, this is the point. All the examples we currently have in (4) of "compounds whose elements themselves contain hyphens or spaces" are compounds which would contain an en dash anyway, according to (5). So either we should find other examples that are not covered by (5), or we accept that this is not a distinct use.--Kotniski (talk) 11:11, 24 June 2011 (UTC)
 * The only type of example that comes to mind (of a compound where we have spaced elements but where we would otherwise use a hyphen) is something like "New Zealand–style clothing". Would we in fact use an en dash there?--Kotniski (talk) 10:33, 26 June 2011 (UTC)
 * CMOS16 gives Chuck Berry–style lyrics, GMAU states that some authors would use a Pulitzer Prize–winning author, and MWM calls for usage in all open compounds, showing Boston–Washington train (apparently they don’t care about the implication of to. Given the extent that we propose to use en dashes, I think we should use them here as well. If we maintain that the en dash doesn’t clarify here, the justification for many of the other becomes weak. It’s hardly that people would not sort it out; as several of us have noted, most publications probably would avoid most uses other than number ranges, either because they don’t care, or because they assume (Chicago, Bill Walsh) that most readers probably won’t see the difference. I assume those of us who do use en dashes do so to provide slight clarification for the readers who do know the difference, and accordingly, if we are going to call for en dashes, we should do so wherever they might help. It’s unlikely that people will read “new (Zealand-style) clothing”, especially with a well-known proper noun, but they may recognize “(New Zealand)-style clothing” a few microseconds earlier. JeffConrad (talk) 21:54, 26 June 2011 (UTC)
 * It's been claimed that this usage of the dash is typically American; if this is the case, hyphens should explicitly be allowed (and possibly the trans-Atlantic difference be mentioned, but that's not vital). ― A. di M.​<i lang="ga" xml:lang="ga">plé​dréachtaí</i> 23:13, 26 June 2011 (UTC)
 * “It’s been claimed . ..” I dunno . .. I read OSM 5.10.9 as allowing, and perhaps preferring, the en dash in this case; I don’t see a short-form alternative. And OSM certainly don’t appear to recommend the hyphen. One gets the feeling from TKE that Fowler would have preferred NewZealand-style clothing, though he conceded the reality of London-New York loan (which is probably less a problem than New York-London loan). But even if usage is strictly American, replacement of a hyphen with an en dash should be allowed in an article in AmE. This of course prompts a question that I think was raised long, long ago in this discussion: if a usage is “typically American” or “typically British”, should usage not match the EngVar of the article? 00:39, 27 June 2011 (UTC)
 * Sure it should, but all other things being equal, a usage which is well-attested in all major standard English varieties is better than one which is only found in the variety of the article (WP:MOS:COMMONALITY). (This is why I usually prefer -ize spelling in BrE.) ― A. di M.​<i lang="ga" xml:lang="ga">plé​dréachtaí</i> 20:47, 27 June 2011 (UTC)
 * In general, I agree, but were we to deprecate or disallow any usage that allegedly is exclusively American or British, we might find rather few uses for en dashes. More important, we’d probably never find agreement. Many Americans, possibly including me, might claim that the spaced en dash is almost exclusively British, and would find support from OUP, but I’m sure examples of use in the US can be found. Perhaps “it’s been claimed” that the en dash in place of a hyphen in open compounds is exclusive to the US, yet OUP seem to support it. If we’re OK with post-Civil War period, we should be fine with New York-London flight. Were we to look at actual usage, we might find that aside from number ranges, use of en dashes is almost exclusive to authors of style guides and those who spend far too much time on Talk:MOS. Incidentally, you’d find much support from OUP for preferring -ize in most cases.
 * Not all of them, only the ones where a completely satisfactory transatlantically accepted alternative exist. ― A. di M.​<i lang="ga" xml:lang="ga">plé​dréachtaí</i> 12:17, 28 June 2011 (UTC)
 * BTW, I'd trim the wording to: “Unspaced en dashes are used, in place of a hyphen, in compounds where at least one of the elements contains itself a hyphen or a space (pre–World War II technologies, ex–prime minister, golf ball–sized object, eka-lead–like properties), but usually not when joining a prefix to a hyphenated element (proto-Indo-European). Hyphens are also possible, especially in British English : pre-World-War-II technologies, ex–prime minister, golf-ball-sized object, eka-lead-like properties. Recasting the phrase may be better style than compounding (e.g. rather than non–foo-baz bar whatchamacallits, consider whatchamacallits other than foo-baz bars.” (Small text denoes stuff I'm unsure whether to include, and the last pair of examples is a placeholder until someone comes up with plausible ones.) ― A. di M.​<i lang="ga" xml:lang="ga">plé​dréachtaí</i> 23:24, 26 June 2011 (UTC)
 * Judging from my American guides, the prefix exception is typically British, so I would not state it as a general case. JeffConrad (talk) 00:39, 27 June 2011 (UTC)
 * I think it would help to maintain parallelism between the dashed examples and the suggested recastings; I had to do a quick double take on first reading it as it stands.
 * That sentence could read, “Recasting the phrase may be better style than compounding (e.g. rather than non–foo-baz bar whatchamacallits, consider whatchamacallits other than foo-baz bars.” ― A. di M.​<i lang="ga" xml:lang="ga">plé​dréachtaí</i> 00:04, 24 June 2011 (UTC)
 * Agree. JeffConrad (talk) 00:38, 24 June 2011 (UTC)
 * I agree on showing the hyphen alternatives; I might also include something like post–Civil War period (to me, it starts getting dicey when the compound is longer than two words). I question ex-prime minister, though—does this suggest that the minister is past her prime? JeffConrad (talk) 22:56, 23 June 2011 (UTC)
 * Agree on swapping (4) and (5).
 * I’d strike “Notice that” in (5) in accord with our normal practice. It’s actually a bit confusing because we haven’t yet shown an example of a hyphenated last name.

Focus
Not sure where to put this, so I'll put it here at the end: if we're going to start discussing whether certain exceptions should be allowable and so on, I think that should happen at the poll page, not here – I envisaged this page as a place to discuss presentation rather than substance. --Kotniski (talk) 07:01, 27 June 2011 (UTC)

What's wrong with this

 * 1) A useful section; emphasis on the primary meaning of the dash is always welcome; but the parenthetical use of dashes is not distinct from the abrupt change in meaning or construction – it's a special case in which the original intent of the sentence is resumed.

4 and 5 ignore the strong disagreement in the poll on those points; sections 5b and 6c. This also ignores the several recommendations to strengthen the recommendation to avoid compounded compounds, including Tony's (section 5c).Septentrionalis PMAnderson 01:49, 13 July 2011 (UTC)

A new draft, at the central forum for MOS
Thanks are due to Kotniski and the others who have worked on this draft. I have opened a new section at WT:MOS for continued development of a draft, as 16 July approaches and we need plenary discussion. See also a summary of the action up till now, in the section that precedes that one ("Dash guidelines: toward a conclusion"). N oetica Tea? 03:17, 13 July 2011 (UTC)

Changes
Are we still considering this draft? PMA has made some edits which I don't feel reflect the consensus from Wikipedia_talk:Manual_of_Style/dash_drafting, to say the least; does it make sense to just have at this draft? What is the point? ErikHaugen (talk | contribs) 04:44, 14 July 2011 (UTC)
 * I agree with you that some of the changes are non-consensus.
 * It may be worth retaining this draft because at least one commenter indicated a preference for it over Noetica’s draft. If this one is retained, it might make sense to strongly discourage any further changes unless it’s designated as the basis for a possible revision to the MOS (I assume the commenter referred to this draft as it was before PMA’s changes). JeffConrad (talk) 04:55, 14 July 2011 (UTC)
 * The wording on pre–World War II used to require the use of a dash. 14 out of the editors polled opposed such a requirement – some opposed a dash altogether; of the 10 who more or less supported it, four did so reluctantly. Changing that is editing in accordance with consensus. Septentrionalis PMAnderson 18:49, 15 July 2011 (UTC)

Opening sentences
Currently reads: I have a slight problem with the firm "do not"... I would suggest something like: My reasoning is simple... We should not firmly say "Do not" when the majority of Wikipedia editors are going to ignore it. To put this bluntly, the majority of editors don't understand (or care) about the difference between hyphens and dashes, or between em and en (or pea, cue, and arrgh) dashes. Even when they do figure out the difference... because keyboards don't contain dash keys, they are going to use approximations, whether we want them to or not. It is futile to tell "don't". We can inform and encourage them to use the correct character, and offer assistance to clean things up when the don't... but we can't make them do so. Blueboar (talk) 14:21, 14 July 2011 (UTC)
 * Two kinds of dashes are used on Wikipedia: en dashes (–) and em dashes (—). For input methods, see Wikipedia:How to make dashes. Do not use substitutes for em or en dashes, such as the combination of two hyphens (--); these were typewriter approximations
 * Two kinds of dashes are used on Wikipedia: en dashes (–) and em dashes (—). Unfortunately, most modern keyboards do not contain keys for these characters, and so they must be created through formatting (see How to make dashes.) A less accurate solution is to temporarily substitute a typewriter approximation such as using the combination of two hyphens (--) while writing your article, and request clean up after the fact.
 * Wikipedia:How to make dashes' main recommendation is to copy and paste. If it's futile to ask someone to copy and paste, then I think it's also futile to ask him to request clean up after the fact. I think over 90% of our editors have experience with copying and pasting in word processing software, so requesting a clean up would be a larger project than copying and pasting a dash. Art LaPella (talk) 22:00, 14 July 2011 (UTC)
 * My point is... the typical editor does not understand (or, to be honest) care enough about using correct style to bother with what is correct... If an editor thinks a dash should be used (of what ever type) most of them will simply type in a "-" found on the top of their key board (which, if I read the guideline correctly is actually a hyphen).  They are not going to bother figuring out how to format something if there is (from their perspective) a perfectly good alternative right there on the top line of their keyboard.  I consider myself a fairly typical editor, and I know this is what I would do, no matter how firmly the MOS tells me not to do it.  At least in my case, I know that I am probably getting it wrong.  Since I do know that, I am quite willing to ask for assistance or clean up after the fact.  Especially if the guideline was polite and asked me to do this.  Ask me to do something (especially if you offer to help) and my reaction is usually positive ("I'll do my best")... order me to do something ("Do this, Don't do that") and my reaction almost always negative ("don't tell me what to do... MOS, we don' need no steeeking MOS...  now fuck off and let me write articles the way I want... and besides, the MOS only a guideline so I don't have to follow it if I choose not to... so there!").  Blueboar (talk) 23:00, 14 July 2011 (UTC)
 * "(especially if you offer to help)" Then maybe your version should, in fact, offer to help: "and request [fill in this blank with something: suggestions include WT:MOS, WP:GOCE, WP:Help desk, Blueboar, ...] to clean up after the fact". Would an editor conclude that "Two kinds of dashes" are to the right of the zero on his keyboard? Oh maybe, if he concluded that the underline is an em dash. "(not a hyphen)" would solve that problem. Yes, your version avoids "Do not", but at the cost of a much longer guideline, and I would feel more insulted being told to request a cleanup when a much simpler procedure exists. Art LaPella (talk) 23:32, 14 July 2011 (UTC)
 * Blueboar, I’m beginning to concede that you’re probably right about the typical editor. As I recall, several of us had no problem with removing the third sentence. The only reason for retaining it might be to guide the editor in using the most appropriate typewriter approximation if she decided that entering actual dashes would be too difficult. But we can’t even seem to find consensus on the appropriate typewriter conventions: every guide I have calls for two hyphens for an em dash and a single hyphen for an en dash (and I learned them using a real typewriter). But I′m not sure most editors are familiar with these approximations—beginning with the 15th ed. in 2003, CMOS no longer covers typewriter-style manuscript preparation. Further complicating things, some have editors have called for automatically replacing double hyphens with en dashes, apparently under the assumption that editors who do not know how to enter dashes will be familiar with the TeX conventions (two hyphens for an en dash, three hyphens for an em dash). I seriously question this, and also wonder how the software would distinguish the TeX afficionados from everyone else. And of course, this assumes that a typical editor will understand when each mark should be used—as I see it, a longer learning curve than that for learning how to enter the characters.
 * So I guess I agree with Art, Dick Lyon, and a few others that the best approach is probably to indicate what’s wanted, guide the editor to the specifics of usage and the methods of entry, hope for the best, and resign ourselves to the need for copy editing. Many experienced professional writers need copy editing, so we should probably not be surprised if it’s needed here. On the positive side: it’s a lot easier to fix dashes than to untoss word salad. JeffConrad (talk) 06:11, 15 July 2011 (UTC)

How about simplifying to:
 * Two kinds of dashes are used on Wikipedia: en dashes (–) and em dashes (—). These do not appear on most keyboards; there are several ways to enter them; see How to make dashes.

MAC fans may prefer many keyboards.

This is all we need to say; describing typewriter approximations only to advise against them (especially when read editors who haven't met them anywhere) is WP:BEANS. Septentrionalis PMAnderson 18:57, 15 July 2011 (UTC)