The headword lines of Ancient Greek entries should show macrons ⟨ ¯ ⟩ and breves ⟨ ˘ ⟩ rather than avoid showing macrons and breves. For instance, the headword line of Εὐριπίδης should show Εὐρῑπῐ́δης.
Option 2:
Macrons: ⟨ ¯ ⟩ and breves: ⟨ ˘ ⟩ shall be used in all applicable instances, where vowel quantity is inferable, in these locations:
Ancient Greek headword lines;
Ancient Greek inflection tables;
Ancient Greek terms in all etymology sections; and,
Lists of (one or more) linked Ancient Greek terms in Related terms, Derived terms, Descendants, Synonyms, etc. sections.
This prescription shall not apply to page titles, quotations, and/or example sentences. Furthermore, this prescription shall not be taken to make the addition of terms without macrons and/or breves by a human a bad edit, but shall be taken to make the removal of macrons and/or breves by a human or a bot (for any reason other than that the information concerning vowel quantity conveyed by those macrons and/or breves be false or uninferrable) a bad edit.
Vote started: 00:00, 5 October 2015 (UTC)
Vote ends: 23:59, 30 November 2015 (UTC)23:59, 7 December 2015 (UTC)
@ObsequiousNewt: We could use transliteration instead of pronunciation to mark length, couldn't we? So even if we do not want to rely on largely non-existent IPA pronunciation sections, and even if we believe they will not be provided any time soon, it should be real easy to extend transliteration with macrons and breves. --Dan Polansky (talk) 06:36, 11 October 2015 (UTC)[reply]
@ObsequiousNewt: Not sure I understand. Do you mean that the reader will see a disconnect between the form and its transliteration (macron in translit not driven by anything in the typography of the form), and will be confused by it? --Dan Polansky (talk) 16:11, 11 October 2015 (UTC)[reply]
@Dan Polansky: Is it intended that transliterations will be added to accompany genitives in nominal declension lines, feminines and neuters in adjectival declension lines, and every form in declension and conjugation tables in order to convey this vowel-length information? — I.S.M.E.T.A.00:10, 14 October 2015 (UTC)[reply]
Oops. @ISMETA: My idea is that transliteration would be provided above all on the headword lines, as it currently is. I don't see the need to provide it anywhere else, but I don't oppose it either right now. Transliteration does not damage the typography of the forms being documented. In Option 1, it is the headword lines that are under discussion, anyway. --Dan Polansky (talk) 15:53, 17 October 2015 (UTC)[reply]
@Dan Polansky: Currently, only the lemma (nominative singular [masculine]; {{grc-verb}} only displays the lemma) is transliterated in headwords. If you oppose macrae and brachiae generally, transliterations would need to be added to accompany genitives in nominal declension lines and to accompany feminines and neuters in adjectival declension lines; otherwise, how would we convey the vowel-length information? Yes, this section only concerns headword lines, but as you wrote in #Oppose option 2 (timestamp: 10:49, 10 October 2015), "[your] preferred option is to have no macrons anywhere for Ancient Greek", so the question of what to do about declension and conjugation tables does need to be addressed. We certainly could add transliterations to the forms in tables; technically speaking, all it requires is the addition of two more parameters per transclusion of {{grc-decl}} and {{grc-adecl}} (though it might be more complicated in the case of {{grc-conj}}). — I.S.M.E.T.A.16:54, 17 October 2015 (UTC)[reply]
The length information would be conveyed in the transliteration in the dedicated entries for the inflected forms. Alternatively, inflection tables could feature transliteration, possibly as a tooltip shown on mouse hovering over. As yet another option, the forms with macrons and breves (not transliterations) could be placed to a tooltip to appear on a mouse hovering over, in inflection tables and elsewhere; thus, the information would be relatively quickly available everywhere, but not without a mouse hovering over the form. --Dan Polansky (talk) 17:29, 17 October 2015 (UTC)[reply]
@Dan Polansky: Entries for Ancient Greek non-lemmata are not customarily created. It would be yet another major change to editing convention. Would you run the bot to create them all? And besides all that, it significantly increases the amount of time it would take a reader to compare forms' vowel lengths. The tooltip idea is the first genuinely acceptable alternative you've suggested; can you flesh out the details for instituting such a system? — I.S.M.E.T.A.19:58, 17 October 2015 (UTC)[reply]
Re: "Entries for Ancient Greek non-lemmata are not customarily created": Not sure what you mean. The general practice in the English Wiktionary is to create entries for inflected forms. For some languages including Czech, the editors decided to focus on lemmas first. No, I would not run a bot to create them. As for tooltips and fleshing out details, I don't see what there is to flesh out: show the macron-free form without mouseover and show the macron-enhanced form with mouseover. As for the technology for tooltips, here is the use of {{comment}}: ἀετός; there does not need to be any underlining. And here's the same thing with {{comment-link}}: ἀετός. --Dan Polansky (talk) 20:11, 17 October 2015 (UTC)[reply]
Information in tooltips should be redundant since it's inaccessible for mobile viewers. Mobile views were apparently 36.47% of total Wikimedia page requests (including bots) in June 2015. Enosh (talk) 13:01, 23 October 2015 (UTC)[reply]
Well, if it could look like that, it would be great, but for me it does not display well. I'm almost certain this is a function of the browser, too, as I have tried manually overriding the font to several otherwise compatible options. Given that this isn't really a problem we can fix, I would defer using combining diacritics where possible. —ObsequiousNewt (εἴρηκα|πεποίηκα) 16:04, 23 October 2015 (UTC)[reply]
Support A good Ancient Greek dictionary should mark vowel length. Vowel length is phonemic, and essential for scanning poetry. Liddell, Scott, and Jones are inconsistent in marking it, but that's a deficiency on their part. (I got very frustrated with them when they didn't mention the long upsilon in ὑμεῖς at all!) No reason for Wiktionary to follow their bad example. As for only marking long vowels in the Latin transliteration or the IPA, that'll just be confusing. As a reader, I would wonder why Wiktionary refuses to write macrons on the Greek when books outside Wiktionary do. It's clunky how the rough breathing in ῡ̔μεῖς displays to the right of the vowel, but phonemic accuracy is worth the cost of typographical ugliness. — Eru·tuon04:49, 18 October 2015 (UTC)[reply]
Re: "books outside Wiktionary do": multiple but not all paper dictionaries do, while publications of actual writings mostly don't. For details, see the talk page of this vote. --Dan Polansky (talk) 05:59, 18 October 2015 (UTC)[reply]
Hmm, I was being quite unclear. I mean some books outside of Wiktionary do: for instance, Hansen and Quinn's Intensive Greek Course, which I learned my Classical Greek from. They mark long vowels with macrons fairly consistently. That probably prepared me for reading Homer later on. Yeah, I haven't seen an edition of a text that uses macrons; my volume of the first twelve books of the Odyssey doesn't. It makes it harder to scan the meter, because you have to guess at which vowels are long and which are short. Anyway, my point is that when books mark vowel length, they mark it on the Greek. It's weird to suggest only marking it on the transliteration. — Eru·tuon21:31, 18 October 2015 (UTC)[reply]
BTW FWIW there's hardly any ugliness for me in ῡ̔μεῖς; the rough breathing is centered over the macron, just as it is in ὑμεῖς. The only difference is that in the former case I see a reversed apostrophe if I zoom in, rather than a small c-like mark, but it's barely noticeable at normal font size. I see the same in Chrome and Safari; this is on Mac OS X 10.9, which may be the underlying reason. Benwing2 (talk) 01:01, 19 October 2015 (UTC)[reply]
@Benwing2: Huh, I'm using Chrome on Windows 10 and the rough breathing displays to the right of the upsilon. Perhaps Apple or a different font is the thing that fixes the problem. It would be nice to figure it out, so that that information could be provided to readers on a help page. — Eru·tuon05:51, 19 October 2015 (UTC)[reply]
@Erutuon, I'm so meta even this acronym: I checked using "Inspect Element" under Chrome; there's a piece of info in the lower right that says "Rendered Font", and it says "Helvetica" for the entire paragraph including ῡ̔μεῖς. I see essentially the same behavior under Safari (also Mac OS X 10.9) so this is probably indeed a Mac thing. BTW I checked to see what the output would be for really weird characters; for the following h1 header, the result is "Georgia" for the 8 glyphs including "Editing ", and "Free Mono" for the 4 Umbrian glyphs:
@Benwing2: Thank you. That doesn't look bad at all. (FWIW, I only see the apostrophe-type psile on my screen.) Could you upload the screenshot of my user-subpage that I linked to, please? I'd like to see how each of the letters is handled, and especially the capitals. — I.S.M.E.T.A.23:30, 21 October 2015 (UTC)[reply]
Sorry, didn't read your previous posting clearly enough. Here it is: screenshot, ISMETA test of macrons/breves with breathing marks, Mac OS X 10.9 Chrome
@Benwing2: That's great; thank you very much. I see that positioning above isn't perfect, but it's a damn sight better than positioning to the right at the top of the line. The capitals are still problematic, because even if it is deemed unnecessary that macrae and brachiae be positioned to the left at the top of the line, the breathing marks really have to appear there. Still, that's an improvement. One last screen-shooting request: I've added a third line of characters to ISMETA/Test (the ones beneath the separator line, which have three diacritics each); would you mind taking a screenshot of those, and then uploading the image herein, please? — I.S.M.E.T.A.12:28, 22 October 2015 (UTC)[reply]
Very sorry to take so long! Here it is: Benwing2 - ISMETA - test Nov 10 If you have any more requests I'll be happy to upload them. Benwing2 (talk) 22:28, 10 November 2015 (UTC)[reply]
@Benwing2: Much appreciated. That's pretty much what I expected it would look like — i.e., pretty damn ugly. Still, at least it's fairly clear what all the diacritics are, so it's a functional improvement, even if it isn't much of an aesthetic one. — I.S.M.E.T.A.03:35, 11 November 2015 (UTC)[reply]
Oppose Superfluous to pronunciation section, and making accents groups less legible, sometimes illegible. Also, macrons and breves could be added to transliteration if the info they provide is so cherished, without typographically damaging the form being documented. For more, see my post at option 2. --Dan Polansky (talk) 07:06, 11 October 2015 (UTC)[reply]
Abstain — Option 2 has the effect of option 1 and more; however, I wish to avoid the legislative irregularity of having both options enacted with overlapping effect. — I.S.M.E.T.A.00:22, 5 October 2015 (UTC)[reply]
In that case, option 2 comes into effect. From the posts to option 2, the closer of the vote will have enough evidence to determine the actual voter stance toward option 1. That is the advantage of actually having comments rather than bare "supports" and "opposes". When you look at it, Wiktionary votes are just great. --Dan Polansky (talk) 12:11, 10 October 2015 (UTC)[reply]
Abstain It should either be option 1 or 2, or neither. Having both of these voted concurrently is like asking whether we should put cheese or Edam on a sandwich. —JohnC501:06, 5 October 2015 (UTC)[reply]
Abstain This vote is superfluous because as the various discussions show, the editors of Ancient Greek reached a consensus on this issue about a year and a half ago. —Aɴɢʀ (talk) 19:03, 5 October 2015 (UTC)[reply]
You know, it probably would look bad to exclude one or two sections, wouldn't it. Very well. I revise my vote to supportfutue id, I have no idea universally. That said, we need to bring up this issue cross-linguistically. —ObsequiousNewt (εἴρηκα|πεποίηκα) 19:50, 6 October 2015 (UTC)[reply]
@Benwing2: I see nothing in the way of rationale. How do you deal with the fact that the combinations with macrons and breves look real bad, sometimes to the point of being illegible, as per the oppose section listing examples of such combinations? --Dan Polansky (talk) 06:27, 17 October 2015 (UTC)[reply]
This already seems to be the consensus. That 's why I think this is a poor vote. In general I think including proper vowel length markings in the Greek is a good idea and the right thing to do. The combinations of accents and macrons look totally fine to me; perhaps this is just how they're displayed in Chrome but in any case we shouldn't compromise correctness just to deal with some browser flaws. Benwing2 (talk) 19:18, 17 October 2015 (UTC)[reply]
Oppose: I would limit it to the closed set of etymology, headword, inflection, derived terms, related terms. Maybe add etymology and descendants sections from other (non-AG) pages into that set. This is to say that I oppose marking length in *nym and see-also sections, and on other pages. (Also in anagrams sections, although Ancient Greek doesn't seem to have any—yet.) My reason for opposing usage in those sections largely is due to the fact that I see no advantage in marking length in those places, and a slight disadvantage in that length marking is less clean-looking. —ObsequiousNewt (εἴρηκα|πεποίηκα) 16:03, 6 October 2015 (UTC)[reply]
Oppose My preferred option is to have no macrons anywhere for Ancient Greek. Since, unlike traditional dictionaries, we have pronunciation sections, the macrons and breves are inessential. They do not appear in the material that we look at for attesting quotations. They do not apply to Koine Greek (see discussion page of this vote). And they make the accent groups above letters less legible and discernible, especially in case of multiple accents: see ᾰ̓ ᾰ̔ ᾰ́ ᾱ̓ ᾱ̔ ᾱ́ Ᾰ̓ Ᾰ̔ Ᾱ̓ Ᾱ̔ ῐ̓ ῐ̔ ῐ́ ῐ̈ ῑ̓ ῑ̔ ῑ́ ῑ̈ Ῐ̓ Ῐ̔ Ῑ̓ Ῑ̔ ῠ̓ ῠ̔ ῠ́ ῡ̓ ῡ̔ ῡ́ ῠ̈ ῡ̈ Ῠ̓ Ῠ̔ Ῡ̓ Ῡ̔, (three diacritics) ᾰ̓́ ᾰ̔́ ᾱ̓́ ᾱ̔́ Ᾰ̓́ Ᾰ̔́ Ᾱ̓́ Ᾱ̔́ ῐ̓́ ῐ̔́ ῑ̓́ ῑ̔́ ῐ̈́ ῑ̈́ Ῐ̓́ Ῐ̔́ Ῑ̓́ Ῑ̔́ ῠ̓́ ῠ̔́ ῡ̓́ ῡ̔́ ῠ̈́ ῡ̈́ Ῠ̓́ Ῠ̔́ Ῡ̓́ Ῡ̔́ kindly collected on the talk page by ObsequiousNewt. But if we must have these macrons and breves at least somewhere, that should be on headword lines and not everywhere, including derived terms, English etymologies, etc. Another alternative is to place macros and breves to romanizations, where they supply the same information without the need to learn IPA, and do not disturb the typography of the form being documented. See also the talk page of the vote for supplementary information. --Dan Polansky (talk) 10:49, 10 October 2015 (UTC)[reply]
See, and after all this time, I'm still swayed by this option, because after all, it does look terrible. The argument that "it should be added to derived terms because it sometimes changes"—well, so does meaning, but we don't typically add that, and Dan's solution of putting it in just the transliteration actually seems to me to be a really good idea. I'd still definitely leave it in the headword and the inflection (the latter of which kind of precludes the LSJ-style of avoiding bad combining diacritics that I would espouse using), and I wouldn't exclude it altogether from unlinked terms in the etymology (e.g. intermediate forms, perhaps where a vowel has been lengthened by some process). Maybe someday if the fonts improve, I would prefer using length marks everywhere, as then the main argument against them is no longer a problem, but, until then... I find myself quite undecided on the whole matter. (Certainly I would prefer this argument in terms of the amount of work it saves me, but that's not a good reason for anything.) —ObsequiousNewt (εἴρηκα|πεποίηκα) 15:41, 10 October 2015 (UTC)[reply]
@ObsequiousNewt: Marking variable vowel lengths in transcriptions only is also problematic. Benwing only fairly recently ran a bot to eliminate a great many manual transcriptions; we would need to put them back, and restart adding them as standard. The main display problems involve capitals and/or breathing marks; length marks combine OK with the oxia and baria on minuscules. ⟨ Ᾰ Ᾱ Ῐ Ῑ Ῠ Ῡ ⟩ already display improperly, anyway — the macra/brachia should be positioned at the top of the line to the left of the letter (like all the other diacritics when they appear on capitals), not above it. When I learn how to design fonts, this is a ubiquitous display problem that I intend to tackle. Until then, there is always the factor that, the more visible a display problem is, the more likely it is that it will be fixed by developers; perhaps our use of all these combining diacritics will accelerate the pertinent improvement of these fonts. — I.S.M.E.T.A.15:48, 11 October 2015 (UTC)[reply]
⟨ Ᾰ Ᾱ Ῐ Ῑ Ῠ Ῡ ⟩ look okay to me, with macrons and breves appearing above the tops of the letters, on an analogy to Czech Á or Č, where I am used to and expect the diactitics to appear above the top lines of letters. It may not be according to someone's standard, but is perfectly legible and good looking. --Dan Polansky (talk) 15:57, 11 October 2015 (UTC)[reply]
Re: "the macra/brachia should be positioned at the top of the line to the left of the letter (like all the other diacritics when they appear on capitals), not above it": Let me list also some other diacritics above capital A, collected from W:Latin-1 Supplement (Unicode block): À, Á, Â, Ã, Ä. On my font and screen, all the diacritics appear above the letter, for non-A letters as well. My printed copy of GÖDEL, ESCHER, BACH has Ö such that the diacritic is above the letter. Therefore, the commented statement seems false. --Dan Polansky (talk) 16:20, 11 October 2015 (UTC)[reply]
He's saying so because the other diacritics are always on the upper left (e.g. Ἆ)—but this is hardly necessary for the macron and breve, which weren't used anyway (true, Woodhouse marks them thusly, but it's not necessary.) —ObsequiousNewt (εἴρηκα|πεποίηκα) 18:05, 11 October 2015 (UTC)[reply]
Abstain This vote is superfluous because as the various discussions show, the editors of Ancient Greek reached a consensus on this issue about a year and a half ago. —Aɴɢʀ (talk) 19:03, 5 October 2015 (UTC)[reply]
The fact that Ancient Greek editors started adding macrons (and sometimes breves) to entries about a year and a half ago, and that other Ancient Greek editors did not revert those changes but started doing the same thing themselves is evidence of consensus. Wiktionary is not a bureaucracy; not everything requires a discussion if everyone finds themselves in agreement with what's going on, and if there is a discussion and consensus emerges from it, no vote is necessary. —Aɴɢʀ (talk) 10:49, 10 October 2015 (UTC)[reply]
Re: "as the various discussions show, the editors of Ancient Greek reached a consensus on this issue about a year and a half ago": You say that various discussions show something; is that true or did you err in this case? --Dan Polansky (talk) 10:51, 10 October 2015 (UTC)[reply]
Abstain Vowel length (i.e., writing macrons and breves) is only sometimes relevant for etymology, declension, and derived terms (etc.). It is relevant when you're writing out the declension of nouns like ἀλήθειᾰ(alḗtheia), where the final alpha is either long or short depending on case and number form; not so much for nouns like ἀετός(aetós), where the long alpha is in the stem. And vowel length is relevant for the etymology section of μήτηρ(mḗtēr), where the Doric form μάτηρ(mátēr) is listed: this Doric form has a long alpha, because it's the original form that was changed by the famous Attic vowel shift of long alpha to eta. The entry currently doesn't give the Doric form with a macron over the alpha, but it should. In short, editors should be allowed to add vowel length to etymologies, declensions, and other areas of an entry where it is relevant. I would support a modified version of Option 2 that allows, not requires, writing vowel length outside of the headword of an entry. — Eru·tuon04:49, 18 October 2015 (UTC)[reply]
@Erutuon: Please n.b. the last sentence of option 2, viz. “…this prescription shall not be taken to make the addition of terms without macrons and/or breves by a human a bad edit, but shall be taken to make the removal of macrons and/or breves by a human or a bot…a bad edit”, which I take to be essentially equivalent to your “allows, not requires” condition. — I.S.M.E.T.A.16:54, 18 October 2015 (UTC)[reply]
@I'm so meta even this acronym: So you're saying Option 2 doesn't require addition of macrons and breves? If so, it's very badly phrased when it says "macrons and breves shall be used in all applicable instances...". Should say "Macrons and breves may be used...". Either that or I don't speak Bureaucratese well enough. — Eru·tuon21:37, 18 October 2015 (UTC)[reply]
@Erutuon: I was trying to write a proposal that was immune from intentional misinterpretation; obviously I failed (it's not even immune from unintentional misinterpretation), but there's no need for snark. Option 2 says that length marks should be used in all those various places, but that it's OK for a human editor to add new content that lacks length marks, but that no one is allowed to remove length marks, unless the length marks convey incorrect or unknowable information. — I.S.M.E.T.A.22:01, 18 October 2015 (UTC)[reply]
@I'm so meta even this acronym: My apologies; I didn't realize you had designed Option 2 (I can't keep track of all the discussion on the talk page) and I could be a little more tactful. Thanks for the explanation of what you intended with the wording. What you say sounds reasonable, but still I'm confused by the wording, so my abstention has to stand. I would support a reworded and clearer version of Option 2 that reflects your intentions as you've explained them here. — Eru·tuon05:48, 19 October 2015 (UTC)[reply]
@Erutuon: No worries. :-) It's somewhat frowned upon to alter vote texts after voting begins. I'd need to obtain unanimous consent from those in favour before doing so. @JohnC5, Metaknowledge, Benwing2: Would you be OK with my rewriting the vote text of option 2 to make its meaning clearer? — I.S.M.E.T.A.14:15, 21 October 2015 (UTC)[reply]
Abstain as not knowledgeable enough about our Ancient Greek entry editing in general, just commenting though to note that at minimum macrons in etymology sections would be welcome. (I am more hesitant about breves — macron/no macron should be legible enough?) -Tropylium (talk) 21:04, 22 November 2015 (UTC)[reply]
Because of the universally acknowledged hash that has been made of this vote (for which I readily take the lion's share of the blame), there will probably need to be another vote to supersede this one in the near future. — I.S.M.E.T.A.02:06, 8 December 2015 (UTC)[reply]
Hardly. There was a large amount of discussion, not all of which was resolved; several options were proposed during the vote, and multiple people expressed complaints that the vote was oddly designed. Personally, I would like to propose a vote designed along the lines of my comments on the talk page, but not until after we're sure we've looked at all the options. Of course, this could also be the sore loser in me talking. —ObsequiousNewt (εἴρηκα|πεποίηκα) 13:58, 8 December 2015 (UTC)[reply]
Support — I.S.M.E.T.A. 12:44, 10 October 2015 (UTC)
Oppose Superfluous to pronunciation section, and making accents groups less legible, sometimes illegible. Also, macrons and breves could be added to transliteration if the info they provide is so cherished, without typographically damaging the form being documented. For more, see my post at option 2. --Dan Polansky (talk) 07:06, 11 October 2015 (UTC)
Abstain — Option 2 has the effect of option 1 and more; however, I wish to avoid the legislative irregularity of having both options enacted with overlapping effect. — I.S.M.E.T.A. 00:22, 5 October 2015 (UTC)
