Wiktionary:Votes/2015-08/Allowing matched-pair entries

Allowing matched-pair entries
Voting on:
 * Allowing entries for matched pairs, like:, [ ] ,  { } , “ ”, ‘ ’, " ", ' ', „ ”, „ “, ” ”, / /, _ _, * *, « », » «, ⌈ ⌉, ⌊ ⌋, 「 」, 『 』, ¡ ! and ¿ ?.

Rationale (edited from the first message in this July 2015 discussion):
 * 1) (repetition) With separate entries ( and ), most definitions tend to be repeated: sometimes, the left side sense is "Begins X" and the right side sense is "Ends X", thus requiring one to check two separate pages to see definitions for the same thing; also, "begins" and "ends" makes it a bit longer to read, especially when these two words are present in almost all senses in the two pages.
 * 2) (consistency) With two almost identical entries (say, { and }), editing one entry often requires editing the other for consistency. One example of inconsistency (although easy to be fixed) is that as of August 27, 2015, { has a sense that } doesn't.
 * 3) (lexical unit) Arguably, since in most senses, you can't use one parenthesis without the other, they are together only one lexical unit.

Disclaimers:
 * This vote does not intend to decide what exactly to do with separate-part entries such as ( or ). One possibility would be deleting all repeated content which can be found on matched-pair entries. Some of these separate-part entries may still have legitimate unpaired uses, such as 1), 2), 3) in numbered lists.
 * This vote does not intend to decide what exactly is the format for the matched-pair entries. At least four different formats have been discussed:, , ( ... ) and (...).
 * Any of the entries used as examples above can undergo RFD and RFV normally, voting here does not mean endorsing any of those entries in particular.

Known issue:
 * Some punctuation marks are used in vertical writing, such as ⏜ and ⏝ (vertical parentheses) and some CJK punctuation. Until further discussion, any vertical punctuation is considered outside the scope of this vote. In other words, this vote does not endorse the creation of matched-pair entries for vertical punctuation. Similarly, pieces of the same symbol such as ⎛, ⎞, ⎜, ⎟, ⎝ and ⎠ (parentheses pieces) are left outside of the scope of this vote as well.

Schedule:
 * Vote started: 00:00, 13 September 2015 (UTC)
 * Vote ends: 23:59, 12 October 2015 (UTC)


 * Vote created: --Daniel Carrero (talk) 14:24, 27 August 2015 (UTC)
 * Discussion:
 * [[Image:Wikt rei-artur3.svg|20px]] Beer parlour/2015/July
 * [[Image:Wikt rei-artur3.svg|20px]] Beer parlour/2015/August
 * [[Image:Wikt rei-artur3.svg|20px]] Wiktionary talk:Votes/2015-08/Allowing matched-pair entries

Support

 * 1)  --Daniel Carrero (talk) 01:00, 13 September 2015 (UTC)
 * Notes:
 * I created all the matched-pair entries which are linked in the "Voting on:" at the start of the vote. In Talk:「 」, said: "The Chinese and Japanese entries shouldn’t have a space. I will rename this article to 「」 if there is no objection." As I replied in the talk page, I have no objection to using the spaceless format 「」 and 『』 to meet the standards of a particular language. That said, the exact format for matched-pair entries is still up for this discussion and I'd still prefer the spaced format for most languages. In any event, specifically, " " looks much better than "".
 * Matched-pairs are something relatively new in the English Wiktionary (started in July 2015) and most other Wiktionaries don't seem to have them, but at least the Kannada Wiktionary has had kn:" " and kn:“ ” since 2011 and the Ido Wiktionary has used io: as a soft redirect to io:parentezi since 2005. --Daniel Carrero (talk) 20:00, 9 October 2015 (UTC)
 * I don't see why the Chinese and Japanese entries shouldn't have a space. These scripts don't normally use spaces, no, but these matched pairs are separated. The space here is not a word separator, so the normal considerations don't apply. A space seems as good a character as any to separate them. —CodeCat 21:00, 9 October 2015 (UTC)
 * That’s because they just don’t use it. See an on-line dictionary for example : かぎかっこ on Goo辞書. — T AKASUGI Shinji (talk) 14:57, 10 October 2015 (UTC)
 * Actually, in that page you linked, the definition starts with the sentence "文章表記中などで用いる 「 」『 』の記号. " There's a space inside of 「 」 and inside of 『 』. --Daniel Carrero (talk) 05:37, 11 October 2015 (UTC)
 * 1)  — Solid rationale. — I.S.M.E.T.A. 14:12, 13 September 2015 (UTC)
 * 2)    &mdash; Saltmarsh συζήτηση-talk 05:21, 14 September 2015 (UTC)
 * 3)  - makes sense to me WurdSnatcher (talk) 02:08, 19 September 2015 (UTC)
 * 4) . I think it's more logical and more maintainable for “ to have as definitions "1. ", "2. " (and then to define those strings as "encloses a quotation in X, Y and Z languags") rather than to define “ as "1. closes a quotation in X, Y and Z languages", "2. opens a quotation in A, B and C languages" and hope the list of definitions and the lists of Xs, Ys and Zs and As, Bs and Cs in the definitions stay synced. - -sche (discuss) 00:33, 21 September 2015 (UTC)
 * 5)  Vorziblix (talk) 23:53, 7 October 2015 (UTC)
 * 6) . These pairs act as a unit, just like, say, the two halves of a circumfix. And we have separate entries for those, too. —CodeCat 20:07, 9 October 2015 (UTC)
 * 7) . We already have circumfixes such as ge- -t. Spacing is a problem because theoretically  the entry name should match the actual use. In English it should be “” without a space while in French it should be « » with a space. For the visibility’s sake, spacing may be allowed. — T AKASUGI Shinji (talk) 01:34, 10 October 2015 (UTC)
 * 8) . -- Romanophile  ♞ (contributions) 00:50, 11 October 2015 (UTC)
 * 9)  but also allow the unpaired entries to continue to exist. Pur ple back pack 89   16:09, 11 October 2015 (UTC)
 * 10) . --Panda10 (talk) 16:16, 12 October 2015 (UTC)
 * 11)  Aryamanarora (talk) 22:57, 12 October 2015 (UTC)

Oppose

 * 1)  primarily because of all of the confusion that this causes when the symbols have meanings outside of the pairing, and also because I don't think there's a systematic way to title these that wouldn't run into other issues. A simpler solution, I think, would be to generically state that the closing symbol closes various sections started with the starting symbol, without saying exactly what that section could entail. For instance, right paren closes left paren, no more no less, listed as a single definition line. (The only other definition line would be where right paren is used independently.) Now the distinct uses of the pairing are left up to left paren to distinguish. DAVilla 05:33, 15 September 2015 (UTC)
 * 2)  The proposed entries are not the thing attested; the space in the middle tries to stand for something else. If we are really obsessed with avoiding redundancy, it would be better, I think, to make soft-redirect senses in each closing bracket to the opening bracket entry. We cannot get rid of the separate bracket items themselves anyway since e.g. ")" is used alone in numbering styles, e.g. "1)". --Dan Polansky (talk) 11:32, 20 September 2015 (UTC)
 * Re.: "it would be better, I think, to make soft-redirect senses in each closing bracket to the opening bracket entry". I think that would be problematic with the multiple varieties of quotation marks. We have 3 entries using guillemets: « », » « and » » (taken from Quotation marks). If we follow that exact format, then « » would be defined at «; » « and » » would be defined at ». See the table in the quotation marks entries (Template:quotation marks) for more varieties of quotation marks. Also, matched-pair titles make the entries clearer, in my opinion: ⌈ ⌉ was designed to be perceived easily as a pair of symbols that ought to be used together, while just ⌈ and ⌉, separately, are less intuitive.
 * I would like the system to work this way, ideally: contains all the senses with parentheses used together, which are the majority of uses; ( and ) contain any individual-character senses if they exist. Regardless of any individual uses, ( and ) contains senses which are soft-redirects with only the modicum of information to point out the existence of, like how Simplified Chinese points to Traditional Chinese and romanization entries point to their original script versions.
 * Re.: "The proposed entries are not the thing attested; the space in the middle tries to stand for something else."
 * I've been using the same format as Category:English circumfixes, with the space in the middle, of which there are three entries: a- -ing, em- -en and en- -en, not counting other languages. Wikipedia also uses this format in the article Bracket, including in the right-floating large punctuation table. Someone could propose using the spaceless format (as in: “”), though I prefer “ ” for clarity of both individual characters. In particular, I've noticed that the straight quotation marks with the space in the middle (" ") look very much better than the possible spaceless variety, which would be too confusing, and its purpose unclear. ("") --Daniel Carrero (talk) 16:10, 29 September 2015 (UTC)
 * 1)  I don't think this actually does reduce maintenance – in fact, it increases it. Each entry for the individual symbols still needs to link to the matched-pair entries (redirects are impossible, for the reasons stated above), and in order to help people find the correct matched pair, the definitions would still have to be eg. "(Standard German) closes « »; (Swiss German) opens » «". It just moves the definitions around a bit. (The situation is even sillier for straight quotes: "(English) opens and closes """?). As an additional pedantry, quotation marks don't always appear as matched pairs (when dialogue spans multiple paragraphs, for instance, there are more opening quotes than closing ones, while in Spanish you have a degree of free choice about how you arrange your inverted punctuation – ¡como?, ¿como!, ¿¡como!?, ¡¿como?! and ¿¿¿como??? are all valid according to Wikipedia – I don't know whether you could also say ¡¿como!?).  Smurrayinchester (talk) 09:46, 25 September 2015 (UTC)
 * (Also, I can't see how we'd handle right-to-left languages properly – we'd need separate pages for the identical looking &rlm; ” “ &rlm; (Arabic quotation marks, opened by ” and closed by “) and “ ” (English quotation marks, opened by “ and closed by ”), for instance, and that fractures definitions even further.) Smurrayinchester (talk) 10:08, 25 September 2015 (UTC)
 * You have a number of interesting points. As the creator of this vote, I've taken my time to think about them, here are my replies.
 * "Each entry for the individual symbols still needs to link to the matched-pair entries (redirects are impossible, for the reasons stated above)". In my opinion, while redirects seem to be possible in a number of cases (from ⌈ and ⌉ to ⌈ ⌉, from ¡ to ¡ !, etc.), I'd aim for consistency and ideally keep all the individual symbols linking to the matched-pair entries, whether or not the symbol stands for something individually, but that's just my opinion and it can be discussed. (That's why I put the disclaimer: This vote does not intend to decide what exactly to do with separate-part entries such as ( or ).)
 * "the definitions would still have to be eg. "(Standard German) closes « »; (Swiss German) opens » «" No, actually not. Apparently these would be your definitions for the individual symbol », but the fact is that the entry as of now contains just the modicum of information necessary for people to find the three entries that use it: « », » », » «. See also " ". The matched-pair entries should explain regional variations and multiple uses, not the individual character entries. Just "Opens X" or "Closes X" is worthless, because we can't see the full picture, especially amidst many languages with their own variations.
 * I insist, see the older translation sections in the ' and the ' entries and tell me if you can make any sense of them, especially for Hungarian, Romanian and Swedish. In addition, there is the fact that in these two entries, “ and ” were not linking to each other properly in a way to easily find each other, in case someone were actually consulting Wiktionary to use a pair of quotation marks in writing. I'd argue that these entries were unhelpful, and most likely understandable only to people who already know how to use the quotation marks in English or other languages.
 * Note that we have ?!, !?, ?? and !!. I don't know if we should have !!!, ??? and perhaps other attestable combinations such as ?!?! or whatever. I think we shouldn't. WT:CFI says nothing about repetitions of punctuation marks, but as for words, it limits us to pleeease and not pleeeease and other longer ones.
 * I take your point about right-to-left languages. But can't “ ” explain the Arabic situation too, since it's written exactly the same and only the direction changes? I've added a usage note in that entry with one example in Arabic. I've used the Translingual section for that usage note, as opposed to creating a separate Arabic section for it, but I guess either would be fine with me until further discussion. --Daniel Carrero (talk) 13:05, 25 September 2015 (UTC)
 * You can just create ” “ other than “ ”. You can use  for the page title. — T AKASUGI Shinji (talk) 14:41, 10 October 2015 (UTC)

Decision

 * Passes: 12‒3‒0 (80%). — I.S.M.E.T.A. 00:03, 13 October 2015 (UTC)