Wikipedia talk:Manual of Style/Archive 19

This archived discussion on quotation marks continues from Wikipedia_talk:Manual_of_Style/Archive_18.

No consensus for straight quotes continued

 * You can enter curly quotes very easily by clicking on the curly quotes in the “Insert:” box on the edit page, or by copying and pasting. This isn’t exactly rocket science.
 * Yes it is ugly, but nothing about the proposed guideline forces anybody to do anything. If you put in straight quotes, someone else for whom it is easy can change them to curly quotes. You're not obligated to do anything. That’s the beauty of the wiki system. Nohat 01:07, 12 August 2005 (UTC)
 * Two things occur to me. First, I really can't tell the difference between the two unless I edit the article, since in the default font for Wikipedia curly quotes look exactly like straight quotes. We might want to consider why we're even bothering if most visitors aren't going to even see the difference.
 * Second, it's very true that special characters have long been used in articles, but mostly they are in contexts where it is obvious what is missing if they don't show up. (For example, phonetic symbols. Half the time I can't see them properly and it's annoying, but they are always labeled so that I can tell what it is that's not showing up and it doesn't interrupt the article.) If quotes don't show up, though, you've got missing character boxes all over the place. -Aranel (" Sarah ") 01:12, 12 August 2005 (UTC)
 * I agree with Sarah. We don't need to introduce those curly quotes. No Account 01:17, 12 August 2005 (UTC)


 * FWIW, I find curly quotes a pain in the butt for no discernable gain. -- Jmabel | Talk 06:07, August 12, 2005 (UTC)


 * I don't care much either way as long as nobody complains when they get straight quotes from my edits. However, I'd like to point out this table which shows what people in other countries consider "typographically correct": de:Anf%C3%BChrungszeichen. I'm afraid we may end up with all kinds of interesting quote characters in the English WP. In case your browser doesn't render it: Dutch, German, and English are all similar but each is different. French, Italian, Danish and others use a different type of quotes, but not all of them use them the same way. Rl 07:16, 12 August 2005 (UTC)

The page Quotation mark has used curlies for a long time, long before we switched to UTF-8. That's a good page for experimentation. I understand that many users don't see a difference between straights and curlies on their screen. Indeed, I am quite sure that Windows in the default font and magnification renders exactly the same bitmap. However, rest assured that the are operating systems (Mac OS X, for example) with other font rendering engines that handle them differently. Or print it. Anyway, I don't think the argument They look the same on my screen holds much water. Not everybody uses Wikipedia in the same way.

Again, the are lots of voice about "bother" or "extra work" or "pain in the butt". I don't understand. If you don't care about curlies then by all means use straight quotes. How can the absence of a prohibition make things more difficult? We are not trying to enforce a No straight quotes policy. We are trying to remove a rule that annoys lots of typographically informed people mightily. If you don't care, then by all means continue to not care.

User:RI makes a good point. The MOS ought to point out that English Wikipedia uses English quotation marks, either straight or curly (my proposal does that), and might want to explicitly prohibit angle quotation marks, Swedish quotation marks, German "low sixes", or quotation dashes. I would support such a rule. Arbor 08:36, 12 August 2005 (UTC)

Tech issues: Back to square one
Everybody who hasn't seen this thread before, please start reading at the beginning, which is May 9, and archived under Wikipedia talk:Manual of Style (Quotation marks and apostrophes). (For example, there is a test page that User:DESiegel was looking for.)

Anticipating the switch to UTF-8, User:Susvolans suggested to drop the curlies verboten rule, because it was based (to a large extent) on a technical problem that would be obsolete when English Wikipedia performed the switch. Many arguments pro et contra were given on that thread, and I haven't seen anobody reach a conclusion different from no consensus. What remains is to decide what to do with that conclusion. I was under the assumption that this observation meant that the rule needs to go, since the mission statement of WP:MOS is "It illustrates standards or conduct that are generally accepted by consensus to apply in many cases". That (and wiki philosophy) closes the case, but we can discuss this. The (grotesque) idea that a consensus statement remains consensus until there is consensus to change it is a Wiki issue that seems to run counter to how we do the rest of Wikipedia, and would benefit from a broader discussion. It has not much to do with curlies.

Anyway the opposition repeats a number of claims that are misunderstandings or fiction. I have replied to them earlier, but let me waste some cyberspace commenting on the curlies break things issue again.

For some reason we now, after three months and a successful migration to Unicode, are back at discussing encoding issues, because of four (4) completely trivial characters. I am annoyed by that discussion, so let me spend some time trying to counter it.

There seems to be a persistent opinion in the curlies verboten camp that "curlies break things." Well, I was pretty aggressive in hunting down such statements in the first two months, asking each and every editor who made such claim to back them up with a concrete example. Nobody has done so. I went so far as to set up a number of test pages on the UTF-8 test wiki in June (before Wikipedia itself switched) to allow skeptics to construct such an example. Nobody replied. The invitation is still open—give me a browser, OS, and possibly an external editor that works with Wikipedia (among other things, it needs to be sufficienlty modern to show very nontrivial CSS, and also needs to honor the HTTP encoding header) but fails with curlies. I'd be really happy to see this setup. Really. We need to fix something very subtle in that case, so we better get it out into the open.

However, I have come to believe that this claim is ficticious. There are thousands and thousand of Unicode characters in use on Wikipedia, including English Wikipedia. I cannot even imagine why the four curlies should be more difficult. (Indeed, they are far less difficult, being included in pretty much every typeface ever designed.)

But the strongest argument in favour of curlies don't cause problems is not the absence of examples to the contrary. Instead, it is the positive evidence of hundreds of thousands of Wikipedia pages in other languages that use curly quotes (and angle quotes and low-nine quotes and whatnot) each and every day without any reported problems I know of. That's very strong evidence in favour of curlies don't cause problems. On the other hand we have User:Jonathunder claiming: "Smart quotes break things and make it harder for other editors. I wish people would understand that".

User:DESiegel is a lot more help: "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 hope you haven't see that on Wikipedia, otherwise inform the tech-wizards with all speed. It sounds as if you have looked at a page that submitted a wrong HTTP header. For example, somebody wrote a page on a Mac (say) and saved it in the native Mac encoding. However, his or her web server sends the page in isolatin1 or in UTF-8. Your browser correctly tries to show the page in whatever encoding the page claims to be in. But it was never saved that way. Your browser works according to spec, the web designer (or the monkey managing his web server) fouled up. There could be other explanations (we who fight this professionally can tell long stories about it). But none of this is relevant for Wikipedia. It is saved in UTF-8, its edit box sends and receives UTF-8, and your browser will understand the header informing it to switch to UTF-8. Trust me, it's going to work. The only possible problem is cutting and pasting, or using external editing software that isn't Unicode aware. But that's going to foul up lots of other characters as well—dashes, funny accents, interwiki links, etc.—so you would break the page anyway, curlies or not. Arbor 08:19, 12 August 2005 (UTC)


 * To summarize:
 * * No one has shown any evidence that curly quotes are broken in any commonly-used browser setup. Therefore the argument that the use of curlies will degrade the user experience for any significant group of users is invalid.
 * * No one is going to be forced to use curly quotes! This means that any arguments of the form “I think they're difficult to enter” or “I think they're a pain”, etc. are not relevant to the matter at hand.
 * * There is no consensus supporting maintenance of the current ban on curly quotes.
 * So how again do we justify the MoS continuing to forbid them? Nohat 08:52, 12 August 2005 (UTC)

Very nice summary, Arbor. I've been away from this thread for a while; I'm astonished it's gone on for this long and (seemingly) gotten this contentious.

I think there's one small point which is responsible for much of the remaining contention: minority users of antique browsers.

Big fat disclaimer: up until just a few weeks ago, I myself was a "minority user of an antique browser" (Netscape 4.7). I sympathize entirely with minority users of antique browsers. I am not trying to disparage minority users of antique browsers. I know exactly how frustrating and demeaning it is to try to access what ought to have been a straightforward web page which ought to have been displayable using any browser at all (even an antique one), to have it display badly due to its seemingly gratuitous use of more modern web features, and to be told that no one's interested in supporting the antique browsers any more and that I really have to upgrade, whether I want to or not, whether it's even possible for me to upgrade or not. (In my case, for reasons I needn't go into here, for a long time it essentially wasn't possible for me to upgrade, until I took time out to do a fair amount of hard work.)

Anyway, if you are using an older browser that doesn't support full Unicode or any of the more exotic non-ASCII characters, you're resigned to the fact that you won't be able to see pages written in Chinese or Japanese or full of mathematics or other special symbols. But since your browser can display those two plain-ASCII characters ' and " just fine, it tends to stick in your craw more when you see boxes or question marks or raw uninterpreted UTF-8 gobbledegook where there are obviously supposed to be quotes or apostrophes. It's easy (for you) to hold the attitude that "I don't understand why people think curly quotes are so important; under the best of circumstances they don't look that different; and they look much worse for me, so I wish everyone would accommodate me and keep using straight quotes everywhere." I think this explains why the argument looms so large for just "the four curlies", as opposed to the "thousands and thousands of [other] Unicode characters in use on Wikipedia".

But then there are the people who honestly feel that curly quotes look significantly better -- and I'm a typography nurd who, even when my browser couldn't display them, strongly agreed with this view (and wished my browser could). So the question is, how do we best accommodate both sets of preferences?

There's one final piece to the argument. Clearly, there is no consensus. Clearly, some people need or prefer the straight quotes, and some people prefer the curlies. In the best of all possible worlds, articles could be stored using curly quotes and rendered for display, if necessary for people using older browsers, using straight quotes instead. But -- and this is the key point -- there's an obvious asymmetry here, because while curly quotes can be straightforwardly transliterated back to straight, the reverse isn't true: if an article is encoded using straight quotes, there's not enough information to reliably convert them to curlies for display by users who prefer them that way. (And of course Help:Special characters has always recognized this asymmetry: "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". [But it also concedes, "...unless proper display on old software is critical", which I guess is why we're still arguing.])

At any rate, this is why I must come down on the side of allowing (but not requiring) directional quotes, i.e. not requiring straight ones, i.e. not disallowing the curlies. Directional quotes contain significant information. The importance of retaining that information outweighs the nuisance caused to users of older browsers which can't display them, especially since (a) there are so few of those users, (b) they really do (hate to say this, but it's true) need to upgrade if at all possible, and (c) there's always the chance that Wikimedia could preemptively transliterate directional quotes back to straight on behalf of those browsers.

Steve Summit 16:04, 13 August 2005 (UTC)


 * As an update, 2726 has since been implemented. This converts characters into entities in the edit box on old browsers. Susvolans (pigs can fly) 16:10, 13 August 2005 (UTC)


 * Indeed, and that's excellent news. There are of course two fairly different issues: (1) display of special characters for anyone who even views a page containing them, and (2) convenience and possible mangling when someone with a non-UTF-aware browser tries to edit a page containing them.  The bugfix affects (2) only, but this is the more urgent issue, since it concerns data fidelity and inadvertent breakage (and of course it transcends the straight-vs-curly debate, too).  So the "use of curly quotes might lead to data corruption during editing" argument ought to be closed, and we're left with (1). Steve Summit 16:55, 13 August 2005 (UTC)


 * Lynx in Latin-1 mode automatically approximates characters when displaying, which covers (1). Macintosh IE, which is the most widely used browser not to have a Unicode-compliant editing box, is similar but handles curly quotes directly. Netscape 4 is now an extreme minority browser and the support issues with it are more severe than character sets, in particular I don’t think Wikipedia’s CSS works round it. Susvolans (pigs can fly) 07:36, 15 August 2005 (UTC)


 * In short, there are technical problems only for a minority that we neglect anyway. I will take this short (but edifying) debate as evidence that the problems anticipated by users like Jonathunder and DESiegel are fictitious. (I also note—not without unbecoming conceit—that my plea to produce concrete evidence for such problems has remained unanswered.) I will revert to the no consensus version, and ask others we re-revert me to be more specific about the technical problems that might be caused by using curlies. We would all very much like to solve them. Arbor 20:11, 15 August 2005 (UTC)

THE ANSWER I think we have the answer then. No matter what is decided, I will continue to use straight quotes. -- Ravenswood 16:00, August 12, 2005 (UTC)


 * Fine. (Of course, you do realise that you are going to burn in Hell for that, and that every time you use a straight quote God kills a kitten.) However, I think we had "the answer" for some time now. That's not the issue anymore. Now we have an editing war on our hands over what seems to be a fundamental wikilosophy issue, and I can see no way of resolving it using the usual guidelines. Anybody with experience in conflict resolution who can tell us how to proceed? Arbor 15:48, 13 August 2005 (UTC)


 * Can't we just have a vote? Choose one of the following: A) Use only straight quotes, B) Use only curly quotes, C) Use whatever suits your fancy, D) Don't even bother being consistant within a single article, E) I don't care, I'm going to do things my way. -- I vote 'E', by the way. -- Ravenswood 07:07, August 14, 2005 (UTC)

Consensus or not
Okay, we're on the brink of an all-out revert war here, so it's time to take a step back. I was about to scold the reverters for repeatedly removing the "no consensus" wording when in fact there is no consensus, but then I went back and counted, and it looks like there's more of a consensus than I'd been asserting there wasn't.

I've tallied, I think, everyone who has contributed to the "Quotation Marks - Why Straight? Change proposed" section on this talk page. ( I have not tallied the responses on Arbor's Wikipedia_talk:Manual_of_Style_%28Quotation_marks_and_apostrophes%29 page yet.) Here are the results:


 * pro curly (4): Nicholas, Arbor, Steve Summit, Susvolans


 * maybe pro curly (5): Michael Z., Factitious, Christoph Päper, Tysto, dab


 * pro straight/anti curly (14): DavidH, Atlant, Jonathunder, jguk, DES, RI, Ravenswood, Antaeus_Feldspar, CDThieme, Satori, Woodstone, Aranel, No Account, Jmabel


 * no opinion (2): Mike, Aya_42

These "pro" and "anti" labels could be misleading, since of course not everyone who is "pro curly" is also "anti straight" (i.e. no one has been suggesting mandating curlies or banning straights). And some people (those I've categorized as "maybe pro curly") didn't necessarily say which they preferred, just that they were in favor of dropping the "only use straight quotes" rule.

But -- sorry, Arbor -- 4 is much less than 14, and even 4+5=9 is still significantly less than 14. (I thought I remembered one or two other people voicing vehement esthetic preferences somewhere in favor of curly quotes, but I don't see their comments now, and they wouldn't change the fact that there do seem to be substantially more people here who are arguing in favor of keeping the "use straight" rule.)

Many of the stated reasons, it's true, are weak or have been discredited:


 * 1. Curlies are hard to enter. [doesn't matter; they're not required]


 * 2. Curlies cause problems. [shouldn't matter; many of the problems are hearsay; few actual first-hand problems described; recent mediawiki bugfix reduces problems even further; problems no worse than for other special characters]

But there are also, I concede, some decent arguments:


 * 3. Curlies don't bring that much benefit for the amount of trouble they can cause; we should keep things simple.


 * 4. The curly problems are worse than for other special characters, because quotes (and apostrophes) appear in virtually all text.

(I'm not saying I agree with 3 and 4, just that they're not as easy to dismiss.)

I'll state for the record that I'm still not at all happy leaving the "always use straight quotes" rule standing, because it's in direct contradiction to Help:Special_characters. (Nor is it possible to change that help page to bring it into line with this MoS.) Furthermore, the rule is also progress-impeding, in that as long as it's in place, the work and experience building that might be necessary in order to relax it won't tend to happen; we could stay stuck in the 1960's flat-ASCII world forever.

Don't get me wrong: "Keep It Simple, Stupid" is a vitally important principle, one I often argue vociferously in favor of. But progress is important, too; the two must be balanced somehow. The extreme case of keeping things simple is Luddism; contrariwise, sometimes we have to do some initially-complicated work, or endure some transitory pain, in order to achieve long-term gain. (And yes, I do think that using directional quotes, if they were easy to enter and free of editing glitches and displayed nicely for everyone, would be progress.)

But for now, I'm going to drop my opposition to the "always use straight quotes" rule, as long as the note about contention and the reference to Help:Special_characters remain in place. I suggest we revisit the issue again in a year or two, because sooner or later I expect the tide will turn. (And the existing wording can probably still be improved.)

In the meantime, I have a couple of questions. Currently, there are many wikipedia articles which do use curly quotes. Would it be right or wrong to scold someone who -- in ignorance or defiance of the MoS -- continues to use curly quotes? Would it be right or wrong for someone to edit an article merely to change its curly quotes to straight ones? Would it be a good idea for someone (perhaps a bot) to ferret them all out and change them all?

Steve Summit (talk) 05:57, 16 August 2005 (UTC)


 * Thank you for your very reasonable assessment of the situation. However, I am concerned that your restriction of determining the consensus based only on the commentors in the "Quotation Marks - Why Straight? Change proposed" was not very accurate, and its results are subsequently possibly misleading. My opinion, for example, was excluded from your tally, even though I expressed it when the matter was first broached in May and also over the past few days.


 * Apologies. While tallying I had a fourth, partially-overlapping category "pro lack of consensus" containing Arbor, Steve Summit, Nohat, and Tysto, and when I omitted that category from the posted results I neglected to move your name to one of the others. Steve Summit (talk) 11:44, 16 August 2005 (UTC)


 * Of course it may be true that Arbor has misrepresented the consensus, but I tend to agree that there is no longer (and perhaps there never was) a consensus to out-and-out ban curly quotes. When only a handful of editors are party to a dispute, any vocal dissent of nontrivial size constitutes a Lack of Consensus.


 * As for points 3 and 4 above, I admit that they are decent arguments; however, they make the assumption that curlies actually cause problems. As you state in point 2, this assertion remains conjecture, so until it can be shown that curlies actually cause problems for a nontrivial number of users, points 3 and 4 remain inapplicable. Ultimately, there are no good arguments for banning curly quotes.


 * What continues to baffle me is why the anti-curly contigent has been able to get away with the continual wholesale reverting of attempts to make the MoS reflect the fact that there is no consensus supporting this rule. The current situation where the curly ban remains but there is a small note indicating that the policy is disputed is only barely tolerable. The fundamental concept behind Wikipedia is to continually modify pages so they they better reflect the consensus. Wikipedia practice is to not revert when you disagree, but to try to engage in compromise and continually edit the verbiage so that all parties are in agreement. The fact that those who are opposed to curlies have continuously and fervently reverted without any attempt at compromise reflects poorly on their character. I hope that Ummit's gentlemanly display of willingness to compromise will be followed by others, and that future attempts to modify the wording to reflect consensus will be met with team spirit rather than curmudgeonly reversion. We are perfectly capable of continually editing pages in the main namespace to satisfy all reasonable parties; I see no reason why the same the same should not be true for the MoS. Nohat 07:10, 16 August 2005 (UTC)


 * I've gone ahead and made my own tally:

Maintain ban:
 * 1) DavidH
 * 2) Atlant
 * 3) Jonathunder
 * 4) Jguk
 * 5) DES
 * 6) Rl
 * 7) Ravenswood
 * 8) Antaeus_Feldspar
 * 9) CDThieme
 * 10) Woodstone
 * 11) Aranel
 * 12) No_Account
 * 13) Jmabel
 * 14) dab

Lift ban:
 * 1) Nicholas
 * 2) Michael Z.
 * 3) Factitious
 * 4) Steve Summit
 * 5) Christoph Päper
 * 6) Arbor
 * 7) Nohat
 * 8) Tysto
 * 9) Susvolans

No clear opinion:
 * 1) Satori
 * 2) Mike
 * 3) Aya_42


 * These numbers give 9–14 in favor of lifting the ban. 61% does not make any kind of consensus. There is quite clearly a Lack of Consensus. Nohat 08:45, 16 August 2005 (UTC)


 * Maintain ban. It isn't just the quotes, but along with it comes apostrophes using the right curly single quote.  Of course, apostrophes don't have a direction to worry about.  They would also make it difficult for someone wanting to expand a contraction, when the browser's find on this page function won't find it if you enter the wrong one, and in my default configuration (and many other people's), I don't see the difference.  I don't even think curly quotes look better.  Most of the time they don't look worse either, except for that `really ugly' left single quote when people think that using the keyboard &amp;#96; (&#96;) with the keyboard apostrophe  &amp;#39; (') is a good way to simulate curly quotes.  Gene Nygaard 09:54, 16 August 2005 (UTC)

Comment. I am not sure tallying a lot of opinions that were not given in form of a vote is a valid way of counting. For example, I understood Ravenswood's comments in the sense that he wouldn't oppose lifting the ban, as long as he could continue to use straight quotes. That would put him in "my" camp. (I may be wrong about Ravenswood, but that's exactly my point: we cannot second-guess people's opinions in this way.) But I don't like voting about such things in the first place.

Another comment: Above, it says "Of course it may be true that Arbor has misrepresented the consensus". Let me just point out that I never have made any claims about "the consensus", so I would be surprised I have misrepresented anything. (If that was the case, I apologize, but re-reading my own comments I really cannot see if I was unclear about this—I am actually quite surprised that there are this many curly-fans in the first place!) Instead, my argument goes as follows: The old rule was based on a technological argument that is now obsolete. This means the rule needs to be re-evaluated, especially since it contradicts Help: Special characters, which prefers curlies for reasons of data integrity. Clearly there is now no consensus for the old rule, and I don't need to count to 50% to make that observation. I cannot understand the mindset behind editors who insist on having the MoS claim the opposite: that there should be a consensus to forbid curlies. There isn't even a consensus in Wikipedia's own guidelines. Wikipedia isn't a democracy in the first place, but I certainly don't need a majority to remove the false claim that there is consensus behind "straight quotes only". Arbor 16:06, 16 August 2005 (UTC)

I've not commented on this issue yet, though I've followed it. I'm not sure if I'm "enough" of an editor to affect this decision. I would forstall considering any lack of comments on either side as a lack of support. I'm strongly in favor of curly quotes, but I haven't chimed in yet because the pro-curly camp has argued so persuasively that I didn't feel I had anything to add. It is possible that those in the pro-straight camp feel the same way about their own champions.

From what I can see, each side has valid reasons for preferring their own quotes. Further, I have no difficulty imagining that there are large numbers of people on either side. I don't think we need an up-or-down vote to realize there is no concensus on standardizing on one quote or the other. My reading of the arguments still makes me think that there are no compelling arguments for forbidding either camp's preferred method.

The revert war on the MOS page is silly. Is there some higher authority than random Wiki users to whom we can appeal to avoid the revert war? Because we can't even seem to get a consensus on a consensus... Mikix 00:34, 21 August 2005 (UTC)

For all of those people arguing to allow curly quotes due to aesthetics: How does Wikipedia look better if some articles have curly quotes and some have straight quotes? Isn't consistency an aesthetic value as well? (Rant: The thing that annoys me most about the MoS is inconsistency. When I edit Wikipedia, I just want some guidance on how to make everything look the same ... I almost don't care what that guidance is. Some articles have american spelling, some have english; some articles use an extra comma in lists, some don't; some use metric, some use imperial; now you want some articles to use curly quotes and some not to?  Why have an MoS at all?) Chuck 15:52, 23 August 2005 (UTC)


 * I know I haven't been involved with this discussion prior, but I agree with Chuck. Saying "you can use either" defeats the purpose of having a Manual of Style.  You can't use "no-one's going to be forced" as an argument for promotion, because that's not style anymore, that's disuniformity.  My personal opinion is that straight quotes should be used, but I defer this to my plea that we don't have a "guidance" to use either-or.  Nothing looks worse than scattered bits and pieces all over tha place.  (My short rant about style is on my user page) Neonumbers 11:51, 25 August 2005 (UTC)


 * The need for consistency across Wikipedia is not a particularly persuasive argument. For example, we will never have style guidance saying that people should use the word "start" and avoid "begin" for consistency. Many things can (and should) be left to the writer's preference. There should be a demonstrable need for any guidance to appear in the MoS. Nohat 09:06, 27 August 2005 (UTC)
 * First, if you don't care about consistency, then why are you bothering even posting in this section.


 * I am posting here because I am trying to combat rampant and unnecessary instruction creep.


 * Second, no one (except you) has discussed in any way creating MoS guidance where there is no demonstrable need. I (nor anyone else) has suggested that we just start coming up with random rules. We are talking about a single conflict (curly vs straight quotes) where there is a demonstrable need to create guidance. The evidence for demonstrable need is the existence of this discussion. Chuck 03:54, 28 August 2005 (UTC)


 * If every time there was a discussion about whether something should be done in one way or another, the outcome was yet another policy page, then Wikipedia would have more policy than actual content. That kind of bureaucratic morass would grind productivity to a halt as everyone cited various and conflicting policies that support their contentions. The fewer policies and guidelines there are, the easier it will be for people to create content and quit debating stylistic details. Nohat 20:35, 28 August 2005 (UTC)


 * Incidentally, I disagree: "...demonstrable need to create guidance" is false. My very first contribution to this debate was a suggestion so simply remove any mention of quotes in the MOS. I hate instruction creep, so I would just remove the current section. This will work, demonstrably: I cannot find any mention of quotation mark style in either French or German Wikipedia (at least when I looked, back in May). Yet both work and look fine, and (as far as I can see) this debate has never been an issue. (As before: If you can find evidence in FR or GE to the contrary, please tell me. I could be simply wrong—stranger things have happened.) So if you ask me, after the UTF-8 switch all straight-vs-curly talk should just be removed from WP:MOS, just as it is going to be removed from help:Special characters. There certainly is the opposite of a demonstrable need for a guideline. Arbor 07:01, 28 August 2005 (UTC)


 * I agree with Arbor. There is no demonstrable need for a rule here, as exemplified by the French and German Wikipedias which get along just fine without one. Both straight and curly quotes have their own advantages and disadvantages. The guideline that articles should be internally consistent still stands, but at this time, since neither straight nor curly quotes have the clear consensus of support so there is no valid justification to prefer one or the other in the MoS. Nohat 20:28, 28 August 2005 (UTC)


 * Do remember that the English Wikipedia is miles ahead of the French and German ones in terms of development.
 * Now, from what I gather, if we take mandate straight vs. mandate curly, there is consensus for straight; if we take maintain ban vs. lift ban there is not. So those who wish the ban be lifted use the lack of consensus as an argument for their minority side.  The thing is, that's not a compromise between the two sides, that's just their side &mdash; not that I wish for a compromise, nor that I consider this paragraph of argument.
 * Anyway, a lack of consistency with these would be very very obvious. There was a small debate a while back over whether or not numbers from 11 to 99 should be spelt out "eleven" to "ninety-nine"; but that's not that obvious, so even though I'd like a guideline on it, without one I could live.  Consistency is a perfectly valid justification for preference.  If it wasn't, the length of this manual would be halved and all the supplementary manuals deleted.
 * The consequences of a mixture of straights and curlies are not insignificant. It is exactly this sort of thing that makes Wikipedia look unrespectable, as if it was written by about twenty teams, not by one. Neonumbers 11:52, 29 August 2005 (UTC)


 * Neo, I appreciate your effort of providing a summary of the status quo, but you are hardly being fair: Now, from what I gather, if we take mandate straight vs. mandate curly, there is consensus for straight Absolutely not. I would much prefer a curlies-only Wikipedia, just like French and German are, just like the blogosphere is, and just like every printed encyclopedia (or other publication) is. I just happen to think it's a hopeless case on EN without Mediawiki support, which is why I am not pressing that angle. There is also no consensus for straights in Wikipedia's own documentation, see help: Special characters, and the historical consensus for straights on this page was based on a now obsolete technical difficulty.
 * Look, if I was in a more confrontational mood, I could easily read help:Special characters as prescribing the use of curly quotes “to maintain data integrity”, and of WP:MOS to basically agree, but to stick with straights due to technical reasons in the latin-1 days. Therefore, the argument would go, the change to UTF-8 incurs, automatically, a switch to curlies, and Wikipedia-EN would finally join the rest of the Internet to not look like a typewriter. (It's not 1993 anymore, the slogan goes.) No debate—straights are out. Good riddance. This (somewhat rude) approach would have established the symmetric situation you describe between a mandate straights versus mandate curlies camp, and a compromise proposal (allow both, stay consistent with the page) would have been the likely outcome. I just don't happen to be a big fan of such behaviour, and thought the lifting of the no curlies ban would have been less dramatic, comparable to what happened with em and en-dashes when these became feasible. I had no idea people would actually want to forbid others to (1) use typographically correct punctuation and (2) follow help:Special characters. I still don't quite believe it. Arbor 13:34, 29 August 2005 (UTC)


 * I just ran across this little tempest. I think the pro-curly crew need to remember that there are a lot of people out here who aren't using uSoft Word, or another editor that makes it easy to enter them. Also, they are not only a pain in the a%% to enter, but they are obnoxious in edit boxes as well - you see #1234's or something all over the place. (I do cut-n-paste for major edits, but not for medium or small ones, so this is a real issue - and it applied to stuff other people write too, so a "let each use what they want" argument doesn't float for me.)
 * If you want Wikipedia to look "professional", ask the developers to add code to render them correctly (with a suitable escape to handle the tricky cases), and stop making the rest of us miserable.
 * Further more, I concur with those who saw that before we can get rid of the old rule (use straight-only), we need a consensus to change it, and we definitely do not have a consensus to change it. Now can we please remove the "disputed" tag from the article? It's been discussed to death, and it's not flying. Noel (talk) 06:46, 3 September 2005 (UTC)
 * Hi Noel, thank you for your comments. (1) Could you explain to me how the possibility of using curlies makes it harder to continue to use straights? I seem to be too stubborn to understand the problem. Nobody is forced to used curlies, so why am I making everybody miserable? (2) Curlies will not be presented as HTML entities (the obnoxious #1234s, as you call them). The appear in the edit box just as any other character: “”‘’. (Try it.) This is a process that all entities on WP-EN are undergoing since the switch to UTF-8 and may take some time, so if you look really hard you may still find some entities. (3) Does the “conservative” argument apply to help:Special characters as well? For remember that WP currently includes conflicting guidelines for how to handle curlies. WP:MOS bans them, help:Special characters prefers them. Would you think the removal of the disputed tag is a fair representation of the current state of the debate? Arbor 07:50, 3 September 2005 (UTC)

Okay I’m new to the discussion here, but I’d like to throw in some academia here. First off, the term ‘straight quotes’ is a misnomer. In English there is only ‘quotation mark’, and by definition is curly. From the Oxford English Dictionary entry on quote: “ 7. attrib. and Comb., as (sense 3b) quotation-capping (see CAP v.1 5), -monger; (sense 2b) -justifier, -quadrat; quotation-marks, signs used in writing or printing to mark the beginning and end of a quotation; in English, single or double inverted commas (see COMMA 4) and apostrophes are employed”. You’ll note that it specifically mentions inverted commas. Now, if Wikipedia has decided to use a would-be ‘straight comma’, then yes, these ‘straight quotes’ get my vote. However, as I believe we can all agree that a properly written/typed/drawn comma does indeed have a curvature, then there shouldn’t be an issue here: standard directional quotation marks should be just that: standard. As well, I should mention that at my university, I’ve had classes in which a mixing up the uses for the curly and ‘straight’ quotation mark result in an automatic F. The directional variant is a quotation mark, used to quote materials, title poems, isolate words, abbreviate, &c, whilst the ‘straight’ variant is reserved for word repetition betwixt two lines (the ‘ditto mark’). Therefore, since the technology exists to handle such marks, we should be utilising it, rather than ignoring it. Otherwise why did WP bother using anything but straight ASCII (pun perhaps intended)? Matthew Stuckwisch 6:45, 12 September 2005, (CST)
 * I personally prefer to use and read straight quotation marks. Many may argue that it helps differentiate where the quotation started and ended; however, if you actually read what is written then you will immediately know where the quotation starts and ends. Convenience of use should be the utmost priority. I have always found straight quotation marks easier to type. I see no need to go out of the way to make quotation marks curly for no apparent benefit when the time could be spent adding information. megamanXplosion 04:52, 27 September 2005 (UTC)
 * MegamanXplosion, your personal preferences notwithstanding, could you make it clear whether you would want to forbid others from using curlies? Usually, Wikipedia allows people who care about good grammar, proper dashes, good photography, nice tables, &c to apply their skills, and few other editors feel that this attention to whimsical details prevents them from "adding information". So the question is not whether you care about such things (you don't, that's fine) but whether you would prevent others from caring, and for what reason. Or did you simply misunderstand the proposal? Nobody is forcing or even asking you to abandon your straight quotes. Arbor 07:44, 27 September 2005 (UTC)

Occam's Razor
Perhaps we should take a lesson from Occam's Razor and Entia non sunt multiplicanda praeter necessitatem. I really couldn't care less about this frivilous matter, as it makes no difference. The font doesn't distinguish between the two styles. Since the proper special characters are supposed to be used anyway, this rule simply pluralizes the rule. I remember when people got behind causes that mattered. glocks out 17:35, 28 September 2005 (UTC)