Wikipedia talk:Manual of Style/Archive 18

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 15 Archive 16 Archive 17 Archive 18 Archive 19 Archive 20 Archive 25


The second half of this archived discussion on quotation marks is in Wikipedia_talk:Manual_of_Style/Archive_19.

Quotation Marks - Why Straight? Change proposed.

Can someone explain to me why the MoS states that straight quotation marks be used at all? They are clearly typographically incorrect, visually inferior, and the only reason they exist at all is because of ASCII (that sub‐standard blight inflicted upon computing in ages long since past). The MoS gives the reasons for this as for “uniformity and to avoid complications”. Now I must disagree with both of these – articles can uniformly use correct quotes, and doing so introduces no complications whatsoever. So both of the cited reasons are quite clearly invalid. Unless someone can give indisputable reasons for using them, I propose that the quotes section be amended and brought kicking and screaming into the 1980s. Nicholas 8 July 2005 12:10 (UTC)

Don't agree, and I'm a purist about typography, but mostly for printed publications. It's significant that the code you used for a hyphen ("‐") doesn't render correctly in my browser.
There are lots of reasons why using straight quotes is simple and efficient, and mandating anything else can introduce problems. I know a lot of people are bothered by not using true typographic quotes, but it's a non-issue to me and probably many others, for now at least. I would vote no to a change. DavidH July 9, 2005 03:55 (UTC)
It's significant that the code you used for a hyphen ("‐") doesn't render correctly in my browser.
That's the key for me; when I stop seeing quotes that turn into "no such character" glyphs, then I'll be ready to change away from straight quotes. But right now, there are still far too many inter-operability problems between the various browsers, text editors, word processors, mailers, and operating systems to do anything other than follow the KISS principle and stick to boring old straight quotes.
Atlant 8 July 2005 14:00 (UTC)
8208 (U+2010) is an unambiguous hyphen. I guess some OSs or browsers don't support it yet, along with thin spaces and attribution dashes. But the ASCII hyphen, em and en dashes, typographic quotes and apostrophe don't suffer from the same problem. Michael Z. 2005-07-8 14:54 Z
David H, I don't think anyone is proposing mandating typographic quotes. Isn't the proposal to allow them? Michael Z. 2005-07-9 06:46 Z
I agree with Atlant completely on this. Jonathunder 2005 July 8 15:25 (UTC)

Almost all articles use straight quotation marks, they are easy to see, and easiest to write in the edit boxes. Why make life more complicated than this? jguk 8 July 2005 17:37 (UTC)

How are the typographic quotation marks harder to see? Which web browser are you using? Michael Z. 2005-07-8 18:23 Z
By "see" I meant aesthetics - I can see curvy ones on Internet Explorer, they just don't look so good, jguk 8 July 2005 18:34 (UTC)
Actual quotation marks look fine to me — they're closer to what books and other printed texts use. Factitious July 9, 2005 03:12 (UTC)
Ah, but a computer screen is not print. What looks good in print does not necessarily look good on screen and vice versa. Indeed, we'd expect differences in font, font size, etc. between the two media, jguk 07:25, 10 July 2005 (UTC)
Good typography is easier on the reader and looks more professional than poor typography, anywhere. Wikipedia's default design employs fonts optimized for the screen that support the characters in question. I've seen WP on both Mac OS X and Windows XP, and there is no problem resolving typographic quotation marks. If you're still using Windows 98, then typographic display probably isn't your priority. Barring technical display foibles of particular systems or fonts, I can't imagine anyone arguing that typewriter quotes and spaced hyphens are actually better than real quotation marks and dashes.
Editing is still an issue so we can't require anyone to type typographic quotation marks and dashes, but there isn't much of an argument for banning these characters, especially since the English WP already includes hundreds of pages with Cyrillic, Chinese and Thai text. Michael Z. 2005-07-14 03:33 Z

It seems to me that you are all correct. Straight quotes are easier to type in (a key feature for wiki-markup), but are rendered incorrectly. The obvious solution therefore would be for the Mediawiki software to automatically identify quote pairs, and translate them to the correct unicode characters when rendering the page. MS Word already does this, so it can't be impossible. There could also be an option to disable this in the user preferences if you don't like it. The important point is that very few users are going to have keyboards with keys devoted to all the typographic variants of quote marks, so they should not be allowed in the wikitext. - Aya 42 T C 19:47, 18 July 2005 (UTC)

A) Ms-word currently does this very badly, often getting the wrong kind of quote, and there is no easy way to correct such errors. I doubt that any authomatical solution will handle all cases reliably. B) too many bouseres currently in wide use don't render these characters properly, so causing us to use them is actively harmful. If the automatic translation were off by defualt, that might be ok, but problem A remains. I oppose any use of "fancy" quotes at this time. DES 20:04, 20 July 2005 (UTC)
Name just one browser—that’s what you meant by ‘bouseres’, right?—in common use that suffers from such problems. Just one!
Netscape 4.x. You'll say it's not in “common use”, and you're right, but I know several people who still use it for various reasons. In fact, I was using it myself until a week ago, when I finally managed to build my own version of Firefox. (I couldn't use the Firefox “Linux” download, because I run Linux on something other than x86). Steve Summit 00:18, 29 July 2005 (UTC)
Automatic conversion does not work well enough to be used, but, still, curly quotes should not be forbidden. IPA, Cyrillic pose more problems, likewise the dagger (†) used in biographies a lot. Typographic dashes, – and —, are almost equally well supported as curly quotes. Yet only the latter are not allowed by the MoS.
It may be reasonable to use the ASCII substitutes in the article titles, in those rare instances. I would recommend redirects, though.
“If a law is not necessary, it is necessary not to have it.” (I forgot who said that and how exactly. Some late 18th century French or American revolutionist, I believe.)
Christoph Päper 15:49, 26 July 2005 (UTC)

Agreed that automagically changing straight quotes to directional (à la MS Word) is fraught with peril. But explicitly encoding directional quotes and optionally converting them back to straight (for display, if the browser prefers them that way) at some point along the line is an almost ideal solution, as long as they're encoded portably, and this is essentially what is advocated by Help:Special_characters#Typeset-style_Punctuation. It's significant to the debate, I think, that the Help page rather directly contradicts this MoS entry in this regard, in that it essentially condones the directional quotes. I had already (before discovering this discussion) inserted cross-references so that readers will be aware of both pieces of advice. I propose changing the MoS text to

For uniformity and to avoid complications it is best to use straight quotation marks and apostrophes. If you prefer to use directional quotes, make sure they are encoded correctly and portably; see Help:Special_characters#Typeset-style_Punctuation.

Steve Summit 14:25, 22 July 2005 (UTC)

Check out what this edit did to directional quotes for a good argument why nobody should use them :-/. Rl 14:34, 22 July 2005 (UTC)
No question, but I have a hard time saying that no one should use them just because some people can't use them correctly. (And in any case when they tend to get broken is when people blindly cut'n'paste text between edit boxes and M-------t W--d, and they're going to do that -- and mess stuff up -- whether or not this MoS says "use straight quotes".) Steve Summit 14:52, 22 July 2005 (UTC)
Gaahh! You were too subtle. When you pointed to "this edit", I didn't notice it was my edit. Sheesh. Haven't figured out how that happened yet; it might have been a problem cut'n'pasting UTF-8 text to/from an emacs buffer. Thanks to Susvolans for fixing. Steve Summit 23:53, 22 July 2005 (UTC)
Can we put this to a vote? What are the guidelines for having the community vote on something? I think it's been discussed enough that any further discussion would just be more jaw-waving. Ravenswood 15:46, July 25, 2005 (UTC)
No. There would be too many stupid people voting nay, just because they think they had to use curly quotes then, although it’s just about allowing them. Christoph Päper 15:49, 26 July 2005 (UTC)

I started this debate some weeks ago on a separate page. See Wikipedia talk:Manual of Style (Quotation marks and apostrophes). From what I can see, there is absolutely no reason to keep the existing rule about forbidding curlies. The rule itself was adopted long ago without any discussion, and was based on a technological restriction. The restriction is now gone, so the rule needs to be reevalutated. The current debate shows that today there is no consensus at all behind that rule, so if we follow the way we usually adopt in such case, the normative restriction in the MoS has to go the way of the Dodo—same reasoning as with dashes, or UK/US spelling, or whatever. Note also that no other Wikipedia I know of needs a rule about what quotation marks editors should use, and they all work fine. Arbor 15:49, 27 July 2005 (UTC)

But they don't "work fine." There are still computers and webbrowsers in existance which render them incorrectly. Also, it is much easier to get everybody to use straight quotes than it would be to get everybody to use curly quotes. I honestly don't even know how to type in curly quotes. Ravenswood 16:09, July 27, 2005 (UTC)
(1) They (e.g., French and German Wikipedia) certainly "work fine". And quotation marks aren't even mentioned on their MoSs. Uniquely among Wikipedias (that I know of), the anglophone MoS makes quotation marks an issue. It shouldn't. Millions of WP articles (in French, German, and probably many other languages) are proof. They work fine. (Otherwise point me to any debate about their problems. I couldn't find any).
I think this is a key point. Right now we have a contradiction between this page (which says "Use straight quotation marks and apostrophes") and Help:Special characters#Typeset-style Punctuation (which says of the typographical ones that "Since using these characters maintains data integrity even on those machines that may not display them correctly, it should be considered safe to use these unless proper display on old software is critical"). I tried to temporarily fix Help:Special characters (to highlight the ambiguity by linking to Wikipedia:Manual of Style so people would at least be aware of it), but my change got (properly) rolled back, becase Help:Special characters is a pure copy of m:Help:Special characters and isn't supposed to be edited. So my attempt failed; I couldn't highlight the ambiguity in this way; it quietly persists. But since m:Help:Special characters applies to all wikis, why can't it apply to en.wikipedia? Steve Summit 00:47, 29 July 2005 (UTC)
(2) I have challenged the argument that starts "there are webbrowsers" several times, including setting up a page so that people can test their concerns. I will gladly to it again: give me a concrete example. So far, nobody has ever responded. I invite you to be the first. (Remember that your situation must be a browser that is sufficiently advanced to view the rest of WP, and that your editing environment should be something that can be used to edit the other UTF-8 symbols in WP.) (3) Nobody want "everybody to use curly quotes". That's a straw man. If you cannot type them, use a straight quote. Wiki works. Arbor 17:13, 27 July 2005 (UTC)
As for (3), I'm sorry, I thought we were aiming for consistent use. Ravenswood 17:20, July 27, 2005 (UTC)
I don't think we are -- or should be -- aiming for a single mandated (i.e. consistent) use at this time. It is not easy to get people to use straight quotes -- many people use the curlies by default, perhaps because their software does it for them. But obviously it wouldn't be easy (would be much harder) to get everyone to use the curlies. But the curlies are arguably superior, which is why I don't think we should discourage those who prefer (and are able to enter) them from using them. Steve Summit 00:25, 29 July 2005 (UTC)

Proposal by Steve Summit

Clearly we don't have a consensus here, so I think we oughta do the NPOV thang and address both opinions. Here's my attempt. What do you think? (Note that I have also resurrected -- and revised -- the paragraph on Microsoft Word, which User:Crissov had deleted at 15:36 on 22 July 2005.)

Use of quotation marks and apostrophes

Single and double quotation marks (and to some extent apostrophes) can be entered in two different styles: the "straight" or nondirectional versions ' and ", and the "typographical" or directional variants ‘ ’ and “ ”. Both styles are in current use on the English Wikipedia; there is not currently a consensus mandating conversion to a single consistent usage.

For simplicity and to avoid certain complications, many editors of the English Wikipedia prefer to use the straight variants. However, Help:Special characters (which is a copy of the mediawiki-wide m:Help:Special characters) says of the directional quotes that they "maintain data integrity even on those machines that may not display them correctly" and that "it should be considered safe to use [them] unless proper display on old software is critical."

Therefore, our best current advice is similar to that for British versus American spelling: When entering new text, it is permissible to use either quoting style, but you should not edit existing text merely to change from one quoting style to another. When you do enter or edit any directional quotes, be careful to encode them correctly and portably; see Help:Special_characters.

There are two situations in which more definitive advice can be given. The fancier, directional quotes (which are not simple ASCII characters) can cause problems in article titles, since Wikipedia uses exact matches for link text and default searches. So the simpler, straight, ASCII quote characters ' and " should be used in article titles. Also, in contractions and possessives, there is nothing to be gained by using a right-hand directional single quote ’ instead of a plain apostrophe, and in those cases plain apostrophes (which are identical to straight single quotes) are preferred.

Be careful if you are pasting text to or from a word processor such as Microsoft Word, since doing so can introduce unintended or nonportable directional quotes. It is recommended that you turn off the "smart quotes" setting in your word processor (in recent versions of Word it is is in the "Autoedit" settings, called "AutoEdit during typing"). If you do permit your word processor to insert directional quotes, make sure they are rendered in your submitted article as portable HTML entities, not as OS-specific extended characters (see Help:Special_characters).

Do not use acute and grave accents or backticks (´ `) as quote characters.

Steve Summit 01:48, 29 July 2005 (UTC)


Comments

Steve, thank you for doing something constructive. My comments are as follows

  1. I strongly disagree with the part about article titles. These are the most visible straight quotes/apostrophes, and an eyesore to the typographically trained eye. The technological argument doesn't hold water—we have lots and lots of articles with much weirder letters, and the software handles these just fine with redirects. Check Beth (letter) or Paul Erdös. For an overview, see Wikipedia:Naming conventions (Unicode) (draft). Compared to the weird letters that already work, curlies are trivial from a coding perspective. They also appear on almost all fonts. We are already using many, many codepoints outside 7-bit ASCII, none cause us any trouble. The technological frontier of what works and what doesn't currently are "combining diacritical marks" as in This Is Spın̈al Tap, which renders weirdly. But the four curly quote characters are trivial. The MoS article should not imply that there are technological problems with using them in article titles.
  2. I don't think portable HTML entities are the way to go. A better advice would be to ensure that the external editor is set to UTF-8 encoding, which it must be anyway so as not to mangle all the other more-than-seven-bits characters. A non-UTF-8 editing environment will mangle the page, so it should be discouraged (for broader reasons than just entering curlies). An UTF-8 editing environment will encode the curlies properly, and not as a HTML entity. (The same happens for ö å ß and whatnot, which are used on countless WP-en pages, including titles.)
  3. Strongly disagree with the preference for the vertical typewriter apostrophe (ASCII 39) over the typographically correct choice (U+2019). To quote from WP's own article on Apostrophe (mark):

In most cases, the preferred apostrophe character is the punctuation apostrophe (distinguished as typographic, or curly apostrophe).

To repeat a slogan: Wikipedia is Not a Typewriter. Arbor 08:50, 29 July 2005 (UTC)

Thanks. Good points. Quick comments on your comments: (1) I have no strong feelings about article titles; I was merely repeating the argument from the original text. (But did you mean to say "The MoS article should not imply that there are technological problems with using them in article titles"?) (2) I'm not quite ready to advocate/mandate UTF-8, because I fear there are still far too many ways to botch it (especially when cut/pasting to/from external editors). I suspect some wikimedia fixes are appropriate here, to detect/correct the circumstances in which the common problems occur. (3) I in turn feel strongly about "plain" apostrophes, but that's my opinion which I shouldn't have injected so strongly into the proposal. Steve Summit 12:11, 29 July 2005 (UTC)
Also, while addressing all the P's oV, we probably need one more sentence saying "Some/Many editors strongly prefer the visual appearance of the "typographical" quotation marks and note that Wikipedia is Not a Typewriter. Steve Summit 12:22, 29 July 2005 (UTC)

Let's see if I can abstract all of this into a slightly edited version of Steve's proposal.

Quotation marks and apostrophes: straight or curly?
Single and double quotation marks and apostrophes can be entered in two different styles: the "straight", "typewriter", or nondirectional versions ' and ", and the "typographical", "curly", or directional variants ‘ ’ and “ ”.
Both styles are in use on the English Wikipedia; there is no consensus mandating conversion to a single consistent usage. Our best current advice is similar to that for British versus American spelling: When entering new text, it is permissible to use either quoting style, but you should not edit existing text merely to change from one quoting style to another. When you do enter or edit any directional quotes, be careful to encode them correctly and portably; see Help:Special_characters.
If you want to enter straight quotes, be careful if you are pasting text to or from a word processor such as Microsoft Word, since doing so can introduce unintended or nonportable directional quotes. It is recommended that you turn off the "smart quotes" setting in your word processor (in recent versions of Word it is is in the "Autoedit" settings, called "AutoEdit during typing").
Do not use acute and grave accents or backticks (´ `) as quote characters.

Note that I have removed quite a bit Steve's proposal, and was probably overzealous in doing so. Please put stuff back in. Especially, I have removed any attempt at explaining the reasons behind preferring one variant or the other; I think the MoS should give advice, not invite debate (that's what the talk page is for). Maybe I'm wrong, in that case Steve had good and short suggestions as to how this could be phrased.

I am very unsure about what advice we want to give about encodings. I think the best idea is just to maintain a link to Help:Special characters, which discusses the issue and would be the best place to keep up to date. I repeat may claim that neither French nor German Wikipedia have anything about this in their MoS, so I boldly assume that there really isn't a problem and the MoS shouldn't scare the user. But I may be underestimating the problem. Arbor 09:25, 1 August 2005 (UTC)

According to Plugwash’s comment at Wikipedia:Village pump (technical)#Lock_out_unsafe_browsers?, there is now a workround to handle non-Unicode-compliant browsers. Susvolans (pigs can fly) 09:45, 1 August 2005 (UTC)
Ah. Good to know. This confirms my opinion that such issues don't belong to the MoS, and certainly not to the section on quotation marks. And the Wikimedia Help:Special_characters seems to need an update badly. Arbor 10:06, 1 August 2005 (UTC)
JFTR, I support Arbor’s version, but IMHO the paragraph on Word is not necessary (any more). Should the section mention or encourage redirects? Christoph Päper 00:05, 5 August 2005 (UTC)

I object to this change in toto -- I think the comments above show no consensus to permit the use of non-straight quote marks, and i think the MoS should continue to recomend that only straight quotation marks be used at this time. I don't see the above discussion as establishing a consensus broad enough to justify a change to an existing MoS provision. Please revert the change recently made to the Mos pending further discussion. DES (talk) 21:31, 7 August 2005 (UTC)

I can see that CDTieme has reverted the change already. Could you or he suggest a method of operation to move this debate along? Its been over two months now since the first post on May 9 by User:Susvolans (see Wikipedia talk:Manual of Style (Quotation marks and apostrophes)), and the debate was announced on this talk page as well as on the Village Pump. The only conclusion I have seen anybody reach is that there is no consensus. We finally have tried to construct a MOS entry to that effect here, and after some polishing and a few extra days of possible feed-back I inserted it on the MOS page. That seems to be the usual way of proceeding. But maybe I read the evidence wrong. Do you think that there is consensus for a straight quotes only policy? Arbor 08:31, 9 August 2005 (UTC)
As I have stated previously on previous topics, Wikipedia guidelines should reflect common practice and consensus, and any guideline that doesn't reflect common practice and consensus must be changed so it does. There does not appear to be any kind of consesus to forbid curly quotes; therefore the section of the MoS that forbids them must be stricken. The "conservative" argument; i.e. that's the way it has been so that's the way it should continue to be, is anathema to the Wiki process. (For the record, I strongly support the use UTF-8 and correct typography). Nohat 09:08, 9 August 2005 (UTC)
I support correct spelling, and that doesn't seem to be a Wikipedia standard. I see relatively little use of curly quotes in articles and small difference at normal type size. On my own website, I use curly quotes only in headings, where I think it's worth the extra effort. I would be more in favor of curly quotes (at least as an option) if they could be added more easily. Can buttons be added to the edit toolbar? I hardly use any of the ones that are there now, but curly quotes and an em-dash would be very useful. --Tysto 13:54, 2005 August 9 (UTC)
I added curly quotes to the "Insert" toolbar that appears at the bottom of edit pages on June 28. They're at the end, between superscript 3 and the Euro sign. Using the "insert" toolbar in most browsers, adding curly quotes is as easy and pointing and clicking. For reference, that template is located at MediaWiki:Copyrightwarning. Nohat 08:17, 11 August 2005 (UTC)
Tysto, thank you for your comments. Ways to enter curly quotes are very much a Mediawiki question and doesn't belong here. Also note (as I have pointed out before) that millions of Wikipedia pages (for example, in Germany and France) uses proper quote marks without software support or outlandish keyboards or any MOS discussion. Also, millions of anglophone Mac users can enter curlies without problems and have done so for decades. Still, before we can politely ask the Mediawiki designers to provide special quote support (maybe only for the Anglosphere? maybe even with new syntax as with dashes? who knows?), the MOS needs to allow using these symbols in the first place. Otherwise the wizards would waste their precious time implementing something that is forbidden to use. I completely agree that software support would be nice, of course! But we cannot expect the tech-gurus to preempt our decision. Ah, and one more thing: on my machine (yes, it's a Mac) the difference between “ and ” and " is quite visible to the trained eye. In headlines and titles it's an eyesore. Arbor 14:17, 9 August 2005 (UTC)
"...millions of Wikipedia pages..."? Aren't there only about 1.5 million total Wikipedia articles? Anyway, I agreed that curly quotes are okay for typographer-types; I'm just saying the great majority of Windows users aren't going to use them if it isn't easy. But do I blame them for not being German? Nein! I blame typographers for not lobbying the keyboard industry to abandon their awkward and outdated layout. --Tysto 18:36, 2005 August 9 (UTC)
Right. I should have said "hundreds of thousands". My bad, thank you for correcting that. As for your points: Windows users (and anybody else) are free to use straight quotes. You seem to be misunderstanding the change. It says no consensus instead of use curly quotes. Arbor 19:09, 9 August 2005 (UTC)
Right. I'm agreeing with the proposal on the table that both are okay, which is the opposite of the original proposal to use curly quotes only. --Tysto 19:35, 2005 August 9 (UTC)
Huh? There was never a proposal to use curly quotes only. (Unless I have missed that.) Could you be more specific? Arbor 07:04, 10 August 2005 (UTC)

I agree with DES here. When wholesale changes to the MOS are made over the strong objections of many and there is not even a majority in favor, the best response may be a revert. And here there are strong objections because of what is lost by these changes. Smart quotes just don't work the same way on my Linux editors the same way they do on Microsoft Windows products. Smart quotes break things and make it harder for other editors. I wish people would understand that. As for the logical quote style, that has been raised and discussed at length. The longstanding guidance is prefered because it preserves exactly what is quoted: no more, no less. Jonathunder 15:18, 2005 August 9 (UTC)

in my view, this is not only a question of typography. It is also a question of parseability (aka 'semantic wikipedia'). I. e. while I hate the curly quotes, I think that quotes, beginning and end, should be formatted so that they are easily recognized by a parser. The ideal thing would be a {{quote|blah blah}} template. Typographical issues then go to the template (they can even be customized via CSS). dab () 18:06, 9 August 2005 (UTC)

C'mon. That cannot have a bearing on the MOS—we don't normally argue that way. However, parsing curly quotes is easy, there are Unicode Character Categories for them, and the parsability of a page certainly improves by distinguishing opening and closing quotes, because these carry semantics (a parser can tell where the quote starts and where it ends). But I don't see this as a relevant argument. Arbor 19:09, 9 August 2005 (UTC)
Then tell me why they break things when I try to use "Find in This Page" in Mozilla Firefox or "Find (on this page)" in IE 6. Since I don't see a difference in my default font size, that is a problem. Redirects can take care of the article title problem, if they are made (and as someone pointed out, when you don't see the difference, you don't stop to think that they need to be made). But they cannot fix the "Find" problems. Gene Nygaard 09:26, 16 August 2005 (UTC)

No consensus for straight quotes

We have debated this for 3 months now. There is obviously no consensus to use straight quotes only (nor, for that matter, to use only curlies). So we have fooled around with a paragraph to reflect this and then copied it to the MOS page. But User:Jonathunder and User:CDThieme have reverted this to the old version with its blanked prohibition against curlies.

Let me assume good faith here and ask both of you:

  1. do you think forbid curlies is consensus?
  2. should the MOS reflect consensus?

I cannnot see how anybody could answer anything else but ‘no’ to the first and ‘yes’ to the second. But clearly I am missing something. Are you making—implicitly or explicitly—an argument ad historiam along the lines of ‘rules persist until there is consensus to abandon them’? From that perspective your behaviour makes sense; but as far as I understand that’s really bad wiki and not how things are normally handled at Wikipedia (or MOS for that matter).

Note that I am not asking a question about straight quotes versus curlies (there has been enough time for that, and I think every argument has been made). I am asking a question about how you think the MOS should work when there is no consensus. Arbor 06:53, 11 August 2005 (UTC)

I have reverted to the non-prohibitory version because no one has answered the points brought up by me earlier and clarified here by Arbor. I expect that anyone who reverts the page will have reasonable answers to these two questions. Nohat 08:03, 11 August 2005 (UTC)
I have reverted because you not only removed additional information not related to the straight-vs.-curly question but because the question itself is asymmetrical. What's the penalty for using straight quotes when curly would work? The typography might be "boring"? What's the penalty for using curly quotes when curly doesn't work? They don't even appear as quotes at all. When there is a conflict, the MOS should reflect utility over appearance -- i.e., making sure it will look acceptable on all systems rather than really fancy on some systems and really messed-up on the others. -- Antaeus Feldspar 11:53, 11 August 2005 (UTC)
My feeling is a) there was a clear guideline for strignt only, and there is no consensus to change this guideline. I think consensus is needed to change the MoS, and that no consensus should mean the atatus quo ante. b) I have seen curly quotes disply as boxes or question marks in mar browser, and i am using IE 6.0.2900, not an unusual setup. I had seen people discuss test pages in this thread -- I have not seen a link to a test page that has curly quotes on it that I can see. c) In articel titles, the exact match rule that wikipedia searches use will mean that the possibility of curly quotes (and of more than one possible quote character, at that) will significantly exacerbate the problem of finding an article. i won't support curly quotes (abostrophes) in articel titles until the softwear will automatically match all quote varients in a search.
For all the above reasons i think the MoS should stay unchanged. DES (talk) 12:12, 11 August 2005 (UTC)

The people who really want those curly smart quotes keep ignoring the fact that not everybody uses the same software they do, and that these abombinable smart quotes are a huge pain for some people. It seems like some people just don't care about making more work for others. But if they keep ignoring the objections and keep trying to force a change through, they will be reverted. CDThieme 13:41, 11 August 2005 (UTC)

OK, then to summarize:
  • In general, it is not necessary for guidelines on the MoS to represent consensus of opinion as well as common practice.
  • In the case that an old guideline loses consensus support, if the users who prefer to keep the old guideline believe that their argument for keeping is more important than the argument of the users who prefer to change the guideline so that it represents the consensus (or lack thereof), then the old guideline should stay even though the guideline no longer represents the consensus.
  • Even though the editors who support not changing the policy have failed to reach a consensus using their arguments, and a significant number of editors continue to disagree with them that the current policy should stay, the fact that the current policy is extant means that they can railroad through keeping their non-consensus-supported policy.
  • When a guideline no longer represents consensus, as long as the people who support keeping the current guideline are vocal and obstinate enough, then there is no need for a guideline to be changed to represent consensus, because the guideline should represent the opinions of the people who support how the guideline appears in the status quo rather than the representing the actual consensus.
I just want to make sure that I understand the line of reasoning. Please clarify if I'm misunderstanding what's going on here. Nohat 17:19, 11 August 2005 (UTC)
  • If there is consensus to change the current guideline, then it should be changed. I do not think such consensus to change has been established as yet.
  • If there is no consensus about the proper guideline, then the status quo ante should remanin unchanged. I think this is the current situation
  • A guideline that will be unworkable for some editors is not preferable to a guideline that will be unaesthetic for some editors.
  • A guideline that makes articles harder to find is a bad idea, in general.
Those are my view on this matter. I might add that I have mentioned this issue on WP:RFC and at the Pump in hopes of getting a broader range of people involved, one way or the other. DES (talk) 17:28, 11 August 2005 (UTC)
Well, as I explained before, I think it is completely un-wiki and unreasonable to support keeping a guideline that doesn't represent consensus simply because the guideline is the status quo. Wikipedia simply doesn't work under the idea that the status quo should remain in general, unless there is overwhelming consensus to change it. In fact I can't think of anything more anti-wiki than that philosophy. Guidelines should be continually changed to represent consensus and common practice. When a guideline doesn't represent consensus or common practice, then it must be changed so it does. It doesn't work the other way around; i.e. that consensus or common practice should be changed to reflect the guideline, yet that seems to be what you and the other obstinate editors are advocating.
As per you other points, they don't really obtain:
  • Remember that you're not arguing against a guideline that requires curly quotes. You're arguing against a guideline that permits curly quotes. No one will be forced to use curly quotes if the guideline is changed to represent the fact that there is no consensus on whether curly quotes should be used.
  • Permitting curly quotes is no more unworkable than permitting any other non Latin-1 characters, but we have plenty of articles that are full of non Latin-1 characters, including ordinary punctuation like em- and en-dashes, as well as IPA and other more exotic Unicode characters, and yet those articles are flourishing.
  • Permitting curly quotes does not make articles harder to find. Creating articles with curly quotes in the title that don't have redirects with straight quotes does, but that's neither here nor there with respect to this policy. There are already plenty of articles with hard-to-type characters in their titles, like Gdańsk, and other guidelines as well as common sense demand that there be redirects at the easy-to-type variety. Curly quotes in titles would be no exception. Nohat 17:49, 11 August 2005 (UTC)
  • If there is no consensus on what the guideline should be, then what should it be? The general rule as I understand it, is that once a policy or gideline gains consensus, and is established, a change requires consensus, and in the absensce of consensus, no change should be made.
  • There is no consensus to permit curly quotes. (Nor to forbid them, I grant.)
  • I find that curly quotes disrupt display significantly more than en and em-dashes, which seem to be rendered corretly for me and I think for most users. I would support a policy to ban all non-latin characters in article titles, but i doubt that would get consensus.
  • "No one will be forced to use curly quotes if the guideline is changed" I grant, but users will be forced to see and perhaps edit articles containing such characters.
  • Granted that articles with titles containing non-standard characters are only harder to find when there is no redirect, allowing such characters increases the chance of this occuring significantly. At least people using diacritical marks generaly know that these are non-standard, and such editors are likely to create proper redirects. Curly single-quotes, used as apostrophes in article titles, I strongly suspect will be seen as "normal " by those whose software generates them automatically, and the proper redirects are far more likely to be ommitted, IMO.
A question to you: When there is an existing guideline or policy, and there is no current consensus on what the guideline or policy should be, should it be changed to a guideline of "there is no guideline" or "there is no consensus on this issue"? Or what?
Don't you see that an argument of "this will make it significantly harder for a sizable number of users" ought to trump an argument of "This will make things look better"?
You argue that curly quotes won't break existing bowsers in any significant way. Put up a test page, please where there is text that uses such quotes, preferably as many versions as you think ought to be permitted in wikipedia, and post a link to it in this thread. Let me and others see just how much of a problem they cause our current setups, and report our results. DES (talk) 18:10, 11 August 2005 (UTC)
I have modified Carlito’s Way (this version in case someone reverts my changes) to use curly quotes for this experiment. Nohat 21:21, 11 August 2005 (UTC)
As for the other issues you bring up:
  • The guideline should be changed if it doesn't represent consesus. I just really don't understand the argument that the fact that a guideline already exists makes it a valid tiebreaker in the case where there isn't consensus. I don't see how the fact that some particular wording is already existing has any inherent value.
I think this is where the miscommunication is happening. Your position, if I understand it, is "there's not consensus on this rule, so it should change". You see editor's choice as a compromise. What others are saying is "there is not consensus on allowing a change that may make Wikipedia less readable, so that change should not be made". Rather than a compromise, they see editor's choice as permitting curly quotes, which some see as problematic, and until a consensus process allays those concerns, the rule should remain unchanged. I don't think its as much a matter of the existing wording having inherent value. Rather, the existing rule was put in for a reason, and until there is consensus that said reason is no longer valid, it shouldn't be changed. -Satori 23:07, 11 August 2005 (UTC)
That seems like a disingenuous application of contradictory logic. The content of Wikipedia guidelines exists because there is a consensus for the guideline to exist the way it is. It doesn’t matter what people’s reasons are for supporting the content: either there is consensus support for a guideline, and the guideline stays, or there is not consensus support, and the guideline goes. The reasons that some people who support keeping a guideline that doesn’t enjoy consensus support are completely irrelevant to whether or not the guideline should stay. It is a completely unworkable policy to take into account the reasons people may or may not support a policy in deciding whether or not the policy should stay. The only metric that matters is whether or not the guideline has consensus support. Right now, the guideline in question does not have consensus support, so it must be removed. If, in the future, it regains consensus support, then it can be re-added, but the situation where a guideline persists even though it doesn’t have consensus support is intolerable and anti-wiki. Nohat 09:01, 12 August 2005 (UTC)
  • If there is no consensus on how something should be handled, then there should be a guideline that indicates otherwise. It seems tantamount to lying to our users if a guideline page says "you should not do X" when in fact there is no consensus that users should not do X.
  • Yes, I agree that "this will make it significantly harder for a sizable number of users" ought to trump an argument of "This will make things look better" but I reject that the former is true, based on the evidence that other special characters have existed in articles for years without significant outcry that results in a consensus that such characters shouldn't be used. Furthermore, this isn't just a question of aesthetics; it's a question of correctness. Using straight quotes to demarcate quoted material constitutes deprecated Unicode usage. It enjoys common usage, sure, but so does the <font> HTML element. Nevertheless, we use <span> here on Wikipedia, not <font>, because the latter is deprecated. Nohat 22:12, 11 August 2005 (UTC)
The version with curly quotes looks exactly the same to me (in IE6, monobook skin). Only if I enlarge the size to "largest" I can see a difference, but not much text anymore. Moreover, the curly quotes then look rather old fashioned to me, and not fitting well with a sans serif font. My conclusion: using curly quotes is a lot of trouble for no benefit. −Woodstone 22:00:24, 2005-08-11 (UTC)
I though this was obvious, but the benefit is correctness, which I thought was one of our goals. Not to mention the fact that they will be required if Wikipedia is ever to exist in printed form, which is also one of our goals, is it not? Furthermore, not everyone uses sans serif fonts in their browser—I certainly don’t. And nobody will be required to use curly quotes! I wish people would stop making arguments about how much trouble curly quotes are to enter because those are arguments against a straw man representation of what this is about. This is about whether they should be forbidden, not whether they should be mandatory. Nohat 22:59, 11 August 2005 (UTC)

Aside from one person who says curly, or smart quotes look better on whatever software system he or she prefers, and one person who is writing a lot about the process of concensus, there already seems to be a consensus of nearly everyone else who has commented or edited this section says the guidance should stay because of how it affects many editors, and how it varies in look from reader to reader. On many systems, to many readers, curly quotes are downright ugly. For many editors, smart quotes are broken. There seems to be a consensus about this, unless consensus means unaminity. I disagree that it does. Jonathunder 18:15, 2005 August 11 (UTC)

OK, first of all, this is how the Carlito's Way article looks on my browser:

They... look like straight quotes anyway.
Second: Suppose I want to edit this article, I take the first sentence:

In 1975, Judge Torres wrote “Carlito’s Way” and its sequel “After Hours”,

...and somehow I cut the quotes off of the movie's title. When I go to put them back, I can't type-in curly quotes, (A point I have made before [1]), so I type them in straight:

In 1975, Judge Torres wrote "Carlito’s Way" and its sequel “After Hours”,

Question: For those of you who can see the difference, is it ugly? Not the straight quotes themselves, but the fact that both types are used in the same article? Because if it is, what you're proposing would require me to replace all the quotes in an article to straight quotes any time I use straight quotes anywhere. Ravenswood 23:10, August 11, 2005 (UTC)

I note that this example conveniently excludes the section of the page that displays the title of the article, which almost certain contains a noticeable curly quote. Nohat 09:01, 12 August 2005 (UTC)

This archived discussion on quotation marks continues on Wikipedia_talk:Manual_of_Style/Archive_19 >>