Wiktionary talk:Votes/pl-2015-09/Using macrons and breves for Ancient Greek in various places

Status quo ante
A long-term practice in the English Wiktionary was not to place macrons in the headword lines of Ancient Greek entries. This was changed by an undiscused bot run, as seen in Εὐριπίδης, on 16 July 2015 (‎). --Dan Polansky (talk) 14:02, 13 September 2015 (UTC)
 * We didn't put them in because technical limitations made it impractical to do so; when those limitations changed (which was quite a while ago now), we switched to using macra and breves, thus your statement of status quo ante is inaccurate. This vote is pointless, and the issue belongs only among Ancient Greek editors, who already have a clear consensus (as you can see from the response in the BP). You are clueless about what has been going on, and making it into a vote will only extend a perfectly functional situation into bureaucracy, merely to get the same result. —Μετάknowledge discuss/deeds 18:51, 14 September 2015 (UTC)
 * I acknowledge that the determination of status quo is probably a bit trickier than I thought. My method of looking at the majority of entries in the mainspace before certain time point can lead to misleading results, in this case. If it is true that most Ancient Greek editors have preferred to put macrons and breves on headword lines and the only reason this was not widespread was lack of effort to make it widespread and/or lack of technology to drastically reduce that effort, then the results are indeed misleading.
 * I propose to run the vote anyway but with the recognition that the status quo ante is either indeterminate or even supporting macrons and breves. Let people supply any evidence they deem worthwhile as for the status quo ante. Thus, I propose to treat the vote as symmetric as for supports and opposes.
 * From what I have seen in Beer parlour, this vote may be an unequivocal pass. Even so, it does not harm to have things clearly on record.
 * One more note. I would like to see a rationale for having these macrons and breves. Someone said they are used in all Ancient Greek dictionaries. Later: see below to check how far macrons are used in all dictionaries, and in what way.
 * Finally, do we want to have these macrons and breves everywhere? Do we want them in example sentences and quotations and do we want them in the etymologies of English entries? Why? Can't we represent the forms in the way in which they are found in the material we are using to attest them? --Dan Polansky (talk) 09:05, 20 September 2015 (UTC)
 * We use them on headword lines, inflection templates, and etymologies when not being lazy. Dan, you seem to have decided that this is something you want to resist despite being completely wrong above (and almost admitting it, too, which is probably a major milestone for you). But the bottom line is that for something that is wholly language-internal, it seems reasonable that the editors who work on that language should decide by consensus, and put that consensus on the about page — and they have done just that! You say that this vote cannot have any bad consequences if it merely upholds existing practice, but it gives Ancient Greek editors another thing to worry about, and thus only achieves minorly pissing them off. There's no point to that. Dan, I urge you to see the writing on the wall, cancel the vote, and get on with your life. —Μετάknowledge discuss/deeds 15:45, 20 September 2015 (UTC)
 * You claim too much and prove too little. The reader will be able to verify, e.g. using a dump, that the use of macrons for Ancient Greek was uncustomary until recently. On your another note, the matter is not language-internal; the matter is about whether we want to present things as they are, or whether we want to present things adorned in various ways. On yet another note, you're far too short on rationale, as are the other editors that seem to have the same stance as you. --Dan Polansky (talk) 18:17, 20 September 2015 (UTC)

Dictionary practice
Let's check dictionary practice as for macrons.

For ἀετός, current Wiktionary entry shows "ᾱ̓ετός" with macron over the 1st alpha.
 * Liddell and Scott online for ἀετός does not show any macron.
 * blueletterbible.org entry for ἀετός shows a scan of Thayer's Greek Lexicon, which does not show any macron over the alpha.
 * Pape: Handwörterbuch der griechischen Sprache for ἀετός does not show any macron over alpha in its online version.

What are the dictionaries that, for ἀετός, show macron over alpha? --Dan Polansky (talk) 09:35, 20 September 2015 (UTC)


 * I feel that this is a wee bit misleading.
 * First of all, The LSJ does list that the first alpha is long (though Perseus's display has never been that good.) LSJ's practice tends to place length information later in the article if an diacritic shares the same vowel in the headword (see the last line of this better digitization). If you look through some other LSJ entries, they represent the macra with some frequency; though they have a nasty habit of leaving off macra in derived terms if they were already listed on the root lemma.
 * The Thayer's, as well as most other Biblical Greek lexica, describes Koine Greek, a stage by which vowel length distinctions were no longer phonemic. These dictionaries do not list vowel length as it is not metrically relevant at that stage. This is similar to how Stelten's Dictionary of ecclesiastical Latin or Niermeyer's Mediae Latinitatis Lexicon Minus do not list Latin macra where L&S, OLD, or Gaffiot would.
 * Another fine dictionary, the DGE, lists the vowel length under “Prosodia”.
 * I hope that helps! — JohnC5 10:07, 20 September 2015 (UTC)
 * DGE does not show a macron in the headword, does it? Their prosodia item, which says "Prosodia: [ᾱ-]", is something like our pronunciation section, right? I see no macrons in the quotations in DGE, either.
 * Can anyone point me to a link to a scan or the like where I can see macrons in headwords themselves, or even in quotations? --Dan Polansky (talk) 10:11, 20 September 2015 (UTC)
 * As for http://stephanus.tlg.uci.edu/lsj/#eid=1874&context=lsj, its headword is "ἀετός"; at the very end of the entry, it says "[ᾱ always.]". --Dan Polansky (talk) 10:12, 20 September 2015 (UTC)
 * As I explained, the LSJ does not place it on the headword if it conflicts with a diacritic. Here's the entry of πατήρ. Such a worrywart you are! — JohnC5 10:17, 20 September 2015 (UTC)
 * Would it be accurate to say that the number of dictionaries that you have shown to have the macron on the headword line is zero? As for "Such a worrywart you are", getting personal much when substance is lacking? --Dan Polansky (talk) 10:19, 20 September 2015 (UTC)
 * It would not be accurate as the link to LSJ's entry for πατήρ in my previous comment has breve smack dab on the second character. I realize now that clearly a breve, that entirely unrelated squiggle used to make things look fancy, has no bearing on this argument, so here is a macron for you. As for my worrywart comment, that was merely a friendly exasperation. I have not made any personal attacks against you or comments about you thus far. One of the great flaws of this project is how angry people get at one another for no reason. I am by no means here to trade invectives with you; though I am exasperated that we must retred ground for which (in my experience and opinion) there will be monolithic consensus. — JohnC5 10:32, 20 September 2015 (UTC)
 * Thank you for providing substance in http://stephanus.tlg.uci.edu/lsj/#eid=72011 for ναός; the headword entry there is "νᾱός". The use of the macron is LSJ for naos is confirmed in archive.org scan. The macron is not used in LSJ for ἀετός as per archive.org, matching what you said above. I checked the scan in blueletterbible.org for Thayer's Greek Lexicon, for ναός, and there is no macron even for ναός. --Dan Polansky (talk) 10:51, 20 September 2015 (UTC)


 * As JohnC5 wrote above (in his post timestamped: 10:07, 20 September 2015), "Thayer's, as well as most other Biblical Greek lexica, describes Koine Greek, a stage by which vowel length distinctions were no longer phonemic. These dictionaries do not list vowel length as it is not metrically relevant at that stage." What the Blue Letter Bible site calls "Thayer's Greek Lexicon" is properly entitled 's Greek–English Lexicon of the New Testament (the NT was written in Koine Greek); as an example in evidence of this, the BLB scan of the entry for matches that on page 422/1 of Archive.org's digital copy of Thayer's Lexicon. As an example of a dictionary that uses macrae and brachiae all over the place, feel free to peruse this digitisation of Woodhouse's English–Greek Dictionary. — I.S.M.E.T.A. 16:07, 20 September 2015 (UTC)


 * Thanks. Is there at least one dictionary that place macron on alpha in ἀετός? Or is the English Wiktionary currently the only such dictionary? --Dan Polansky (talk) 18:52, 20 September 2015 (UTC)


 * See page 258/2, s.v. “Eagle”:
 * Eagle, subs. P. and V. ᾱ̓ετός, ὁ (Plat.). Sea-eagle: Ar. and V. ᾰ̔λιᾱ́ετος, ὁ (Eur., Frag.).
 * — I.S.M.E.T.A. 11:45, 21 September 2015 (UTC)


 * Thanks for the link to Woodhouse's English-Greek Dictionary. I should have been more specific: Is there at least one Ancient Greek dictionary that place macron on alpha in ἀετός in their headword item for ἀετός, which an English-Greek dictionary obviously cannot do, only Greek-English? --Dan Polansky (talk) 08:08, 26 September 2015 (UTC)


 * *sigh*:
 * 1880,, Handwörterbuch der Griechiſchen Sprache, volume I: Griechiſch⸗Deutſches Wörterbuch. Α–Κ, page 43/1, s.v. «ᾱ̓ετός»:
 * ᾱ̓ετός, ὁ, ion. u. poet. αἰετός, w. m. ſ. […]
 * And before you respond that that doesn't count because it's a Greek–German dictionary: that is an illegitimate objection; it's still a dictionary that lists a Greek headword, complete with combined macra and psile. It happens; that has been conclusively demonstrated. — I.S.M.E.T.A. 11:44, 26 September 2015 (UTC)


 * That's fine; thanks. I'd be still interested to see a Greek-English one, but a Greek-German one is an interesting sort of evidence. --Dan Polansky (talk) 11:53, 26 September 2015 (UTC)

Rationale for opposition
I am not going to start the vote yet to enable more discussion before the vote. My rationale for opposing macrons in headword lines: --Dan Polansky (talk) 12:27, 20 September 2015 (UTC)
 * 1. First and foremost, I want our headword lines to reflect what can be seen in the material used for attestation. Addition of macrons violates that requirement.
 * 2. If all dictionaries used macrons despite item 1, I should probably give up. But as per, e.g. Thayer's Greek Lexicon does not seem to use macrons at all, and LSJ seems to use them only in some entries. A popular online edition of LSJ--Perseus--does not use macrons either, so it does not seem critical.
 * 3. I assume that the dictionaries that use macrons do so as a replacement of a pronunciation section. We do have pronunciation sections so we do not need macrons.
 * 4. If the compromise is made that macrons are used on headword lines, then I still think they should not be used anywhere else, in view of item 1: show the real thing. They should not be used in quotations, English etymologies, lists of derived terms, etc. If they are used in Derived terms, the foreign keys there do not match the primary key, the actual headword, which I see as substandard.
 * 5. When placed over letters with other diacritics such as ἀετός vs. ᾱ̓ετός, macrons can be actually harmful since they make it more difficult to see clearly what the other diacritic is; they make the visual parsing made by the human eye and human vision more expensive and error-prone. That could be the reason why LSJ does not place macron there, as per.
 * By this proposed token, macra should be removed from Latin too, which should never, ever happen. We use macra in many dead languages that have stress distinctions regardless of whether they were written down (Many Germanic languages and for Old Prusssian, we even put them in the pagename, despite not having a direct attestation of them).
 * As before, Koine Greek dictionaries do not as it was no longer phonemic nor used in the writing of poetry or determination of the accent location. The Perseus LSJ does use length diacritics (as seen in the trailing underscore in their entry for ναός), they just did a terrible job digitizing the entries (which holds true across the project). The DGE also uses them in its headwords. Both of these dictionaries are inconsistent in their placement because they tend to mark them vowel length of the base lemma but then not on its derivatives.
 * If they are used in place of a pronunciation section, that is perhaps a tertiary role. In many cases the length of a given vowel gives crucial information about the behavior of the movable pitch accent, about how a word will inflect, and even disambiguate different cases. If you are an expert at the morphology and phonology of AG, some of the ending vowel lengths may be implied by the placement of the accent or by the declension statement, but this is merely a short hand used by dictionaries to save timeand space. Pharr and Smyth both use macra throughout because they are grammatically important. In AG poetry there is an almost symbiotic relationship between the metrical scansion and the grammatical parsing, as the length of vowels disambiguates both. Most other dictionaries do not care in the slightest about the pronunciation of each word per se because they lack the space and because it is, at best, a good reconstruction. Normally they will give the vowel lengths and a series of rules for how to pronounce AG at the beginning and that is it. The reason dictionaries given is because they are hugely important to the grammar and poetry.
 * There are a few places where they should always be used.
 * The headword and declension section for the reasons stated above.
 * The etymology sections because the vowel length of AG terms often effects the vowel length of languages that borrow words like Latin. These borrowings provide crucial comparative evidence about the length of different vowels. Of course English does not preserve AG's vowel length distinctions directly, but I could imagine finding pronunciation differences when English borrows from French, which derives from Latin, which borrows from AG.
 * The derived terms should continue the headword for consistency.
 * Quotations indeed should not necessarily have them except when they disambiguate the grammar.
 * This is a fair point and something with which the AG editors have been trying to deal for a bit now. The Unicode implementation it not perfectly suited for this job, but having the information should take precedence over a minor inconvenience to the reader. Fortunately, this problem in the headword is alleviated by the transliteration, which removes any doubt of what the other diacritics represent.
 * — JohnC5 18:23, 20 September 2015 (UTC)
 * Let me note that the headword itself is not the only place where this information can be provided. Macron can be provided in the romanization next to the headword while leaving the headword itself intact. We are not in the dilemma of either providing the information as the macron in the headword or not at all. --Dan Polansky (talk) 18:54, 20 September 2015 (UTC)
 * In truth, we are not in any dilemma at all concerning this matter. Instead, you are creating bureaucratic impediments to something that is already well accepted. This really does have a long history even if you try to ignore the examples and my attestation.
 * Also, as I mentioned before, the new templates, , and all require this vowel length information and will return incorrect forms when it is omitted. That is how important this information is. — JohnC5 19:08, 20 September 2015 (UTC)
 * For discussing what the status quo ante is, we have . But let me address what you say here: I don't know why you mention these templates; they are from April, June and August of 2015. These templates are not even about headword lines; they are about sticking these macrons in inflection templates. The templates are part of the revolution of which I have no evidence that it is supported by the most productive of Ancient Greek contributors such as Atelaes. The templates are not evidence supporting the claim that the macron information is very important; they only show proclivities of those who created and support the templates. The proclivities of some of these editors can be seen: one uses "№" sign in plain writing, utterly unnecessary; another one has the following on their user page: I feel that I am ſhowing great reﬅraint in not rewriting all my noꝛmal entries in Fraktur. : ), which another editor, the one who created these templates, follows with 𝕰𝖍, 𝖏𝖚𝖘𝖙 𝖌𝖎𝖛𝖊 𝖚𝖕 𝖆𝖓𝖉 𝖏𝖔𝖎𝖓 𝖚𝖘 𝖆𝖑𝖗𝖊𝖆𝖉𝖞. This group of editors seem to love fancy typography and what to me seems like excess adornment. The thing in question is not the provision of information; it is aesthetics of information presentation, where my sense of aesthetics is contrary to that of these editors involved in that revolution. If I am alone, so be it, but that's what I'd liketo find out. --Dan Polansky (talk) 08:00, 26 September 2015 (UTC)
 * Rationale № 1 is a very naïve principle. Most Classical stuff will be written entirely in majuscules, in scriptio continua, and without diacritics (which were invented by circa 200 BC) of any kind. Trying to reflect in our headword lines "what can be seen in the material used for attestation" (viz. 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12) is far less simple than you might think. — I.S.M.E.T.A. 13:28, 21 September 2015 (UTC)


 * A fair point. Nonetheless, would you admit that most Ancient Greek writings--not dictionaries and not grammar books but writings--published today are without macrons, and that it is these publications of these writings without macrons that we are looking at when searching for attestation? --Dan Polansky (talk) 08:03, 26 September 2015 (UTC)


 * As far as I'm aware, the status of macrae in Ancient Greek is similar to the status of macra in Latin. I had a more fundamental point in mind, however: There is no feasible way "to present things as they are" (#Status quo ante: 18:17, 20 September 2015) in our choices of headwords. Even if you're just talking about modern publications (and I would challenge the notion that we should prioritise such derivative works over contemporary inscriptions etc.), I frequently find ϑ used instead of θ and it is very common for ϰ to be preferred to κ; furthermore, consider what's written in Citations:sigma, and the example of that here. I'm afraid that this is an insolubly complex issue. — I.S.M.E.T.A. 20:13, 26 September 2015 (UTC)
 * We can at least try to approach the materials we use to search for attestation. If almost none of that material uses macrons, then my item 1 would lead us to avoid macrons. There may be some issues in trying to achieve faithfulness as you've mentioned, but it really is not all-or-nothing, either perfect faithfulness to the most original materials or a free-wheeling abandonment of concern for materials used to source attestation. --Dan Polansky (talk) 05:57, 27 September 2015 (UTC)

Why two formats, an entry one and a display one
This is a question for those who support placing macrons everywhere, in headword lines, derived terms, example sentences, attesting quotations, English entry etymologies, etc., but maybe also for those who want to see macrons in many but not all of the listed locations.

Why do we have two formats, one without macrons for entries, and another one for what we display? If we actually want to have macrons everywhere, why should this distinction exist? --Dan Polansky (talk) 07:45, 26 September 2015 (UTC)
 * Are you only asking this about Ancient Greek? Or are you wondering the same thing for other languages like Old English and Latin, and the Eastern Slavic languages and Bulgarian with respect to stress marks, and Lithuanian and Serbo-Croatian with respect to accent marks? —Aɴɢʀ (talk) 08:14, 26 September 2015 (UTC)
 * In principle, what underlies the question applies to the other languages as well. The question is easily answered if the additional adornment is constrained to headword lines, and is not used in example sentences, derived terms, etc., as has until recently been the case for multiple of these languages, and for some still is the case. But the recent tendencies to stick that adornment everywhere lead to that question that I am asking, without an answer so far. --Dan Polansky (talk) 08:37, 26 September 2015 (UTC)
 * If the use of display-only diacritics has increased recently, it's because Module:languages makes it much easier to use them than it used to be. Before the module had the ability to strip diacritics automatically, writing "amīcus" and linking to amicus required heavy use of the pipe trick:, which was a pain, and especially difficult in inflection tables. Now that the module strips display-only diacritics, they're much easier to use, so people use them more. That's the only reason why they're used now more than they used to be. As for why we use them at all, they make words easier for learners to read and pronounce correctly. —Aɴɢʀ (talk) 08:46, 26 September 2015 (UTC)
 * That does not really answer my question. I'll ask again: if the supermajority of editors wants to see macrons everywhere, why are the macrons not part of the page name in the wiki database? Since then, you don't need any modules or any such mechanism at all; you just use that which is the page name everywhere, including macrons. Or, a slightly different question, what is the benefit of not having the macrons as part of the page name in the wiki database? --Dan Polansky (talk) 09:16, 26 September 2015 (UTC)
 * I guess because the display-only diacritics aren't usually shown except in pedagogical or linguistic material. You'll see them in textbooks, grammar books, and dictionaries, but not in editions of the works of Aristophanes, Ovid, and Alfred the Great (for the ancient languages) or in newspapers or novels (for the modern languages). They're there for the benefit of learners and linguists; native speakers don't need them. —Aɴɢʀ (talk) 10:03, 26 September 2015 (UTC)
 * Great. Now, why should display-only diacritics be used anywhere except in the headword lines given that, as you say, they are not used in the editions of actual writing? --Dan Polansky (talk) 10:06, 26 September 2015 (UTC)
 * Because they make words easier for learners to read and pronounce correctly, even on other pages than the word's own entry. —Aɴɢʀ (talk) 10:57, 26 September 2015 (UTC)
 * Dan, Dan, Dan. I'm not sure you're dreaming big enough! Latvian and Lithuanian (and many of the other Balto-Slavic family members) have the accentation diacritics in their headwords that do are not used in normal writing. For that matter Modern Hebrew and Arabic have their pesky niqqud and ʾiʿrāb, which no one writes with anymore. I assume we should remove them too? I have often noticed that the use of masculine, feminine, and neuter in many languages for the names of genders (which are never directly attested in the text) often does not correlate to the proper sexual anatomy of the particular word in question and are merely vestiges of Latin and Greek studies. I'm pretty sure we should rename all of them to non-case, non-number agreeing declensional classes A, B, and C. Shall I come up with some other perfectly functional and expected linguistic scholarly traditions that we should remove just to please you? If I logged onto Wiktionary, went to the entry for amicus and there were no macra, I would not trust Wiktionary as a source. It would look like an unprofessional project written by non-experts. — JohnC5 16:26, 26 September 2015 (UTC)
 * In all of the above, you seem to be saying that tradition asks us to use macrons. I pointed out that the dictionaries in that tradition do not have dedicated pronunciation sections. But now to the question in the headword of this section: if we want these pesky macrons everywhere, why don't we put them to the entry form itself? --Dan Polansky (talk) 06:12, 27 September 2015 (UTC)
 * An answer by I.S.M.E.T.A. is at, and starts with "The biggest problem with including combining diacritics in page titles themselves is that it will make keyboard entry of search terms virtually impossible". --Dan Polansky (talk) 07:04, 4 October 2015 (UTC)

Advantage over pronunciation section
Why is the pronunciation section not good enough for the purpose; why do we need macrons in addition? Let me note that the dictionaries linked from do not seem to contain any pronunciation markup or notation, so their use of macrons, when they use them, is their sole vehicle of delivery of that information. --Dan Polansky (talk) 10:13, 26 September 2015 (UTC)
 * Because not everyone can read IPA, and because people who use Wiktionary in conjunction with other learners' materials have come to expect them. If I'm learning Latin and my grammar book says amīcus and my textbook says amīcus and my paper dictionary says amīcus, I'd be confused if the Wiktionary entry said amicus on the headword line, even if the pronunciation section says, and even if my OCT or Loeb version of Vergil says "amicus", and even if I quite deliberately searched for  in the Wiktionary search box. —Aɴɢʀ (talk) 11:02, 26 September 2015 (UTC)
 * You mean that not everyone can learn that ː is a prolongation mark in IPA?
 * And you mean that the expectation of users of these learning materials trump the expectation of readers of the material from which we pick attesting quotations? --Dan Polansky (talk) 11:15, 26 September 2015 (UTC)
 * Everyone can learn it, but not everyone has learned it. The readers of the material from which we pick attesting quotations were learners themselves once, and probably continue to use grammar books and dictionaries that use macrons, so even they will expect to see them. —Aɴɢʀ (talk) 11:20, 26 September 2015 (UTC)
 * Would you agree that learning ː is very easy and that it applies to a variety of languages, so it is probably worth it? --Dan Polansky (talk) 11:23, 26 September 2015 (UTC)
 * I wish everyone would learn IPA in primary school, but until that day comes, we can't assume our readers know it. That's why for modern languages we supplement our IPA information with audio recordings whenever possible. —Aɴɢʀ (talk) 11:25, 26 September 2015 (UTC)
 * That's a red herring. We are not talking learning IPA; we are talking learing ː. In a dictionary that has no macrons at all, those interested in wovel length will quickly pick it up. Again, talking about ː and not complete IPA, as specifically mentioned in my question to prevent evasion (a prevention which failed). --Dan Polansky (talk) 11:30, 26 September 2015 (UTC)
 * It's not a red herring, because people who don't know IPA won't be looking at the pronunciation section to begin with. Also, as you found when you created User:Ultimateria/en-needing-ipa a few weeks ago, most entries don't even have pronunciation information. That's as true for other languages as it is for English. —Aɴɢʀ (talk) 11:55, 26 September 2015 (UTC)
 * I don't see why they should not look at the pronunciation section only since they do not know IPA. In fact, when the section is there, it takes a prominent place of the page so it is actually hard to have no glimpse at it. Furthermore, how would they know that our pronunciation sections use IPA if they don't look at them? --Dan Polansky (talk) 12:06, 26 September 2015 (UTC)
 * I think you underestimate how complicated the IPA can be for nonusers. Macra, on the other hand, are very well understood within the community of Latin users.
 * Also, our pronunciation sections are by no means compendious. There are many different academic pronunciation systems that make use of macra and are not in line with our IPA templates. The headword informs many different scholarly traditions without pigeonholing them into one reconstructed pronunciation. — JohnC5 16:26, 26 September 2015 (UTC)
 * I see no reason to think that someone new to Latin understands what these macrons stand for and why the headwords show something that is not seen in publications of Ancient Greek or Latin writing. That same person new to Latin should understand that a section labeled Pronunciation stands a good chance of containing pronunciation. I acknowledge that it is far easier to add macrons than it is to create complete IPA pronunciations, so macrons are a practical way of quickly rolling out a class of information for a large number of entries; that is a real advantage. --Dan Polansky (talk) 06:40, 27 September 2015 (UTC)

Koine Greek
It was pointed out above that "Koine Greek dictionaries do not [use macrons] as it was no longer phonemic nor used in the writing of poetry or determination of the accent location." Assuming that our Ancient Greek sections also cover Koine Greek, why should their headword lines show something that is not applicable to Koine Greek? --Dan Polansky (talk) 11:18, 26 September 2015 (UTC)
 * Maybe for words that only came into existence after contrastive vowel length was lost, they shouldn't. But for Koine words that were already in use in older stages of Greek, the macrons are there to provide information about those older stages. —Aɴɢʀ (talk) 11:22, 26 September 2015 (UTC)
 * But they are inapplicable to the newer stage, where the headword stands for all stages, I suppose. --Dan Polansky (talk) 11:24, 26 September 2015 (UTC)
 * People interested in newer stages can ignore them if they're present, but people interested in older stages won't know where they go if they're absent. —Aɴɢʀ (talk) 11:26, 26 September 2015 (UTC)
 * To pronunciation section, where this type of information generally belongs? How hard it can be? --Dan Polansky (talk) 11:28, 26 September 2015 (UTC)
 * See the previous thread. —Aɴɢʀ (talk) 11:59, 26 September 2015 (UTC)
 * Ok, "won't know where they go if they're absent" is just bollocks as I hope close to anyone can see. A minimum level of intelligence is required; users who do not seek pronunciation information in pronunciation section cannot be helped. --Dan Polansky (talk) 12:07, 26 September 2015 (UTC)
 * And you have not addressed the point that you are placing on a shared headword something that is not universally applicable to things covered by that headword, as far as I am concerned. --Dan Polansky (talk) 12:09, 26 September 2015 (UTC)
 * Sometimes it is not easy to tell whether a word is Koine or not. Scholarly tradition also still uses the spiritus asper and lenis in Koine transcription even though it wasn't pronounced during that period. The system we are using has been around for quite a long time, and it is honestly arrogant of you to think that you know better, especially since you don't appear to have any stake in any of these languages. — JohnC5 16:26, 26 September 2015 (UTC)
 * It's not about a word being Koine or not. It's about placing macrons to headwords even when the headwords represent Koine phase as well. I mean the Koine phase for all those words in the Thayer's dictionary, which you said describes Koine Greek. To be even more explicit, presumably, most of the words used before the Koine phase did not stop being used in the Koine phase as well, right? Yet you tell me that the Koine phase is one "by which vowel length distinctions were no longer phonemic." So again, it seems that the macrons seem inapplicable to Koine phase; again, not to particular Koine words but to the complete phase of all words. --Dan Polansky (talk) 06:07, 27 September 2015 (UTC)


 * As JohnC5 wrote ("Scholarly tradition also still uses the spiritus asper and lenis in Koine transcription even though it wasn't pronounced during that period."), the psile ⟨ ᾿ ⟩ and dasia ⟨ ῾ ⟩ are also "inapplicable to [the] Koine phase". Would you have us omit them, too, and reduce them to the presence or absence of a leading in 5th-C.-BC Attic and  in 1st-C.-BC Egyptian IPA transcriptions? — I.S.M.E.T.A. 08:10, 27 September 2015 (UTC)


 * @ISMETA: Now that's a good point, one that I missed; your "psile ⟨ ᾿ ⟩ and dasia ⟨ ῾ ⟩" was helpful in that regard. Omitting psile and dasia would be contrary to most publications of writings, right? --Dan Polansky (talk) 08:17, 27 September 2015 (UTC)


 * Not necessarily. See this. — I.S.M.E.T.A. 08:32, 27 September 2015 (UTC)


 * @ISMETA: Thanks. Does it mean that my statement including "most" is incorrect? Since "not necessarily" does not really contradict "most". But thanks again; the trouble with "most" statements is that they are so much harder to refute. --Dan Polansky (talk) 08:44, 27 September 2015 (UTC)


 * I'm not sure. I linked to Matt. 1 because the Greek New Testament is very widely printed, and the text's diacritics are often omitted as unnecessary and/or anachronistic (see, for example, this and this for rationales for that practice). It may well be, if reproductions of early Greek Christian materials are included, that unaccented text makes up more of the grc corpus than accented text. — I.S.M.E.T.A. 01:54, 29 September 2015 (UTC)

Change to vote proposal
On account of the fairly apparent consensus displayed herein (pace the vote's creator), I propose to extend the prescription of this proposal to include brachiae and to their use in all Ancient Greek headwords and inflection tables (assuming technical practicality), to Ancient Greek terms in all etymology sections, and to lists of linked Ancient Greek terms in Related terms, Derived terms, Descendants, Synonyms, etc. sections. I propose that the prescription not apply to quotations and example sentences; furthermore, the prescription shall not be taken to make the addition of terms without macrae and/or brachiae by a human a bad edit, but shall be taken to make the removal of macrae and/or brachiae by a human or a bot (for any reason other than that the information conveyed by those macrae and/or brachiae be false) a bad edit. — I.S.M.E.T.A. 08:30, 27 September 2015 (UTC)


 * I weakly object. I think it will be relatively easy to find support for headword lines, but a bit harder to find support for inflection tables, derived terms and the like. I propose that you create a separate vote for your proposal, or a separate section of this vote so we get Support proposal 1, Oppose proposal 1, Abstain proposal 1, Support proposal 2, etc. But again, a separate vote could be better since the page name of the vote is already very clear.
 * Extending to brachiae, while still being constrained to headword lines, is another matter. But maybe you should first clarify what brachiae are: brachiae is a redlink, and brachia takes me to brachium, which does not help. As for macrae, what's that, macrons? . --Dan Polansky (talk) 08:54, 27 September 2015 (UTC)


 * I've now added my proposal to the vote text. Per your suggestion, I made it "option 2", rather than replacing what you wrote. I understand your point about the vote title, but I concluded that it would be best not to dissipate the focus that this issue has received at its current location. — I.S.M.E.T.A. 00:33, 4 October 2015 (UTC)


 * Fair enough. I have modified the vote to use common terms "macrons" and "breves". This is a public vote, and must be as easy to understand for as many people as possible. --Dan Polansky (talk) 05:06, 4 October 2015 (UTC)
 * I have renamed the vote so the title covers your proposal as well. --Dan Polansky (talk) 05:11, 4 October 2015 (UTC)


 * I have accepted some of your changes, but I have rejected others. — I.S.M.E.T.A. 17:39, 4 October 2015 (UTC)

The vote seems... poorly designed at worst, and inconclusive at best. I would like to move that we close it without consensus and start over, if it's going to be rewritten. And as far as I can tell the options on the table—which should really just be multiple options of one vote—are (with their chief proponents): An alternative would be to have two votes that say, roughly, thus e.g. Polansky would vote against both; I would vote for the first but against the second, and Acronym would vote for both. It's a weird setup, but has the slight advantage of allowing Polansky and Acronym to support option 3 as a secondary option. (It also neglects option 2, but I don't think option 2 has any serious supporters, except possibly Dan Polansky.) —ObsequiousNewt (εἴρηκα|πεποίηκα) 15:40, 30 November 2015 (UTC)
 * 1) no length markings anywhere, except pronunciation (Polansky)
 * 2) length markings only in transliteration (idea which came up in discussion, but I don't think it has support?)
 * 3) length markings in headword and inflection only (myself)
 * 4) length markings everywhere but quotations and the page title (Acronym)
 * 1) Length marks should be used in the headword and inflection.
 * 2) Length marks should be used in all sections excepting quotations and page titles.

Comments from the newt
I actually am kind of thankful to Dan for bringing this up, as I had been considering bringing it up for a while myself. After endless self-debate and experimentation I still couldn't come up with a consensus.

The problem, for me, with showing macrons basically boils down to "Unicode combining forms are unreadable". The next best argument is "too much information for the newcomer, which is weak. Representation is hardly a valid argument, as there was no consistent representation of Ancient Greek; the representation of ΔΕΧΣΙΟΝΗΟΜΟΝΗΥΙΟΝΑΡΕϜΙΛΥΚΟΙΟ that looks like "δεξιὸν ὦμον υἱὸν Ἀρηϊλύκοιο" is entirely a result of academic consensus (and even that disagrees on tremata among other things)... but macra and breves alike are symbols I have never seen outside of the dictionary except in a few transcriptions of papyri. The point being that it's safe to say that, at the very least, quotations should not include length marks.

But the key phrase there is "outside of the dictionary", and, as we all know, Wiktionary is a dictionary. I don't think that anyone disagrees that length marks should be included; the question is merely how. To which I answer that there should be at least once a word with diacritics on it. It's simply the easiest way to read it. It's probably best that this is the headword, but I could be convinced otherwise. The pronunciation section alone won't cut it, though, it's cumbersome to read even if you do know IPA, and, as someone pointed out, /dekʰsion/ isn't as accurate as δεξῐόν. Similarly, I'd put it in the inflected forms, if only for the reason that the inflected forms themselves are usually going to involve length distinction (most notably the first declension). And of course it belongs in the etymology section, for reasons described above.

That said, should they be anywhere else? The current policy (which, by the way, I definitely remember discussing, but I can't find where the discussion was) is to put length marks everywhere except the title, but the policy with most other languages, and the prior policy, was to put length marks only in the places I just mentioned. Dan seems to have accidentally posed the interesting question—why is there a different standard for the page title than for the rest of the page? Perhaps it'd be best to put length marks everywhere, including the page title—but there are two problems with this.

One of these is the problem of combining diacritics. Liddell and Scott faced the same problem we face, which is that it's hard to put length marks on a vowel with diacritics already on it. Their problem was probably with a printing press, ours is with Unicode, suffice it to say that we have a problem. That said, that problem is going to exist until someone solves it, and moreover it's going to be our problem if we put the length anywhere (unless we try for LSJ's solution, but I feel like that'd actually be clunkier.)

The other one is hidden quantities. A large portion of vowels we don't actually know the length of, because they're hidden before two consonants and there are no accent rules to let us know their length. The policy for these is simply to represent them with a bare vowel, which is why (unlike Latin, though I'm not really sure why) breves are used as well as macra. Not that this really inhibits the universal use of length marks, but it's worth pointing out.

So, in my opinion, I'd either keep things as they are, with length marks being placed everywhere except the page title—although I think it'd be better to actually include the page title if that's the idea—or, the minimalist approach, to limit it to headword, inflection, and etymology. —ObsequiousNewt (εἴρηκα|πεποίηκα) 00:58, 29 September 2015 (UTC)


 * The biggest problem with including combining diacritics in page titles themselves is that it will make keyboard entry of search terms virtually impossible (unless you're up for learning a bunch of <kbd class="keyboard-key nowrap" style="border: 1px solid #aaa; -moz-border-radius: 0.2em; -webkit-border-radius: 0.2em; border-radius: 0.2em; -moz-box-shadow: 0.1em 0.2em 0.2em #ddd; -webkit-box-shadow: 0.1em 0.2em 0.2em #ddd; box-shadow: 0.1em 0.2em 0.2em #ddd; background-color: #f9f9f9; background-image: -moz-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: -o-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: -webkit-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: linear-gradient(, #eee, #f9f9f9, #eee); padding: 0.1em 0.3em; font-family: inherit; font-size: 0.85em;">Alt codes). As things currently stand, if I want to search for a term in polytonic Greek, all I need to do is press <kbd class="keyboard-key nowrap" style="border: 1px solid #aaa; -moz-border-radius: 0.2em; -webkit-border-radius: 0.2em; border-radius: 0.2em; -moz-box-shadow: 0.1em 0.2em 0.2em #ddd; -webkit-box-shadow: 0.1em 0.2em 0.2em #ddd; box-shadow: 0.1em 0.2em 0.2em #ddd; background-color: #f9f9f9; background-image: -moz-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: -o-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: -webkit-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: linear-gradient(, #eee, #f9f9f9, #eee); padding: 0.1em 0.3em; font-family: inherit; font-size: 0.85em;">⊞ Win <kbd class="keyboard-key nowrap" style="border: 1px solid #aaa; -moz-border-radius: 0.2em; -webkit-border-radius: 0.2em; border-radius: 0.2em; -moz-box-shadow: 0.1em 0.2em 0.2em #ddd; -webkit-box-shadow: 0.1em 0.2em 0.2em #ddd; box-shadow: 0.1em 0.2em 0.2em #ddd; background-color: #f9f9f9; background-image: -moz-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: -o-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: -webkit-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: linear-gradient(, #eee, #f9f9f9, #eee); padding: 0.1em 0.3em; font-family: inherit; font-size: 0.85em;">Space <kbd class="keyboard-key nowrap" style="border: 1px solid #aaa; -moz-border-radius: 0.2em; -webkit-border-radius: 0.2em; border-radius: 0.2em; -moz-box-shadow: 0.1em 0.2em 0.2em #ddd; -webkit-box-shadow: 0.1em 0.2em 0.2em #ddd; box-shadow: 0.1em 0.2em 0.2em #ddd; background-color: #f9f9f9; background-image: -moz-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: -o-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: -webkit-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: linear-gradient(, #eee, #f9f9f9, #eee); padding: 0.1em 0.3em; font-family: inherit; font-size: 0.85em;">Space and then use the Greek Polytonic keyboard (GPK) to input the necessary characters. The GPK does have key combinations for inputting vowels with macra or breves (viz., for ᾱ and ᾰ, for example, <kbd class="keyboard-key nowrap" style="border: 1px solid #aaa; -moz-border-radius: 0.2em; -webkit-border-radius: 0.2em; border-radius: 0.2em; -moz-box-shadow: 0.1em 0.2em 0.2em #ddd; -webkit-box-shadow: 0.1em 0.2em 0.2em #ddd; box-shadow: 0.1em 0.2em 0.2em #ddd; background-color: #f9f9f9; background-image: -moz-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: -o-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: -webkit-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: linear-gradient(, #eee, #f9f9f9, #eee); padding: 0.1em 0.3em; font-family: inherit; font-size: 0.85em;">- <kbd class="keyboard-key nowrap" style="border: 1px solid #aaa; -moz-border-radius: 0.2em; -webkit-border-radius: 0.2em; border-radius: 0.2em; -moz-box-shadow: 0.1em 0.2em 0.2em #ddd; -webkit-box-shadow: 0.1em 0.2em 0.2em #ddd; box-shadow: 0.1em 0.2em 0.2em #ddd; background-color: #f9f9f9; background-image: -moz-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: -o-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: -webkit-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: linear-gradient(, #eee, #f9f9f9, #eee); padding: 0.1em 0.3em; font-family: inherit; font-size: 0.85em;">A and <kbd class="keyboard-key nowrap" style="border: 1px solid #aaa; -moz-border-radius: 0.2em; -webkit-border-radius: 0.2em; border-radius: 0.2em; -moz-box-shadow: 0.1em 0.2em 0.2em #ddd; -webkit-box-shadow: 0.1em 0.2em 0.2em #ddd; box-shadow: 0.1em 0.2em 0.2em #ddd; background-color: #f9f9f9; background-image: -moz-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: -o-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: -webkit-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: linear-gradient(, #eee, #f9f9f9, #eee); padding: 0.1em 0.3em; font-family: inherit; font-size: 0.85em;">⇧ Shift +<kbd class="keyboard-key nowrap" style="border: 1px solid #aaa; -moz-border-radius: 0.2em; -webkit-border-radius: 0.2em; border-radius: 0.2em; -moz-box-shadow: 0.1em 0.2em 0.2em #ddd; -webkit-box-shadow: 0.1em 0.2em 0.2em #ddd; box-shadow: 0.1em 0.2em 0.2em #ddd; background-color: #f9f9f9; background-image: -moz-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: -o-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: -webkit-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: linear-gradient(, #eee, #f9f9f9, #eee); padding: 0.1em 0.3em; font-family: inherit; font-size: 0.85em;">- <kbd class="keyboard-key nowrap" style="border: 1px solid #aaa; -moz-border-radius: 0.2em; -webkit-border-radius: 0.2em; border-radius: 0.2em; -moz-box-shadow: 0.1em 0.2em 0.2em #ddd; -webkit-box-shadow: 0.1em 0.2em 0.2em #ddd; box-shadow: 0.1em 0.2em 0.2em #ddd; background-color: #f9f9f9; background-image: -moz-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: -o-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: -webkit-linear-gradient(top, #eee, #f9f9f9, #eee); background-image: linear-gradient(, #eee, #f9f9f9, #eee); padding: 0.1em 0.3em; font-family: inherit; font-size: 0.85em;">A, respectively), but none for combining those length marks with marks for breathing and/or those for pitch. For Latin, of course, the increase in searching difficulty that would result from adding macra to its lemmata's page titles would be very much greater. As for the considerations on the opposite end, I suggest that there is some value in being able to discover, at a glance, to what degree vowel quantities are shared by a term's relations and derivations. — I.S.M.E.T.A. 23:49, 3 October 2015 (UTC)


 * @User:ObsequiousNewt: Thank you for your input. Can you please clarify what brachea are? Are these macrons, breves and anything else? ( does not help; nor does web search).
 * Furthermore, as for "current policy [...] is to put length marks everywhere except the title": Where is this policy? You mean some people's practice rather than policy? Or you mean some people's edits to WT:AGRC, which is a think tank, not policy? --Dan Polansky (talk) 07:19, 3 October 2015 (UTC)
 * Later: brachea are probably breves. The would-be term probably derives from βραχεῖα, a form of βραχύς (short, brief, ...); English "breve" derives from Latin brevis (brief, short, ...). --Dan Polansky (talk) 07:11, 4 October 2015 (UTC)


 * I'll be honest, I'm not sure where I got that word. Obviously it's from the plural form βραχέα (not βραχεῖα, which is the feminine form) but I have no idea who actually uses it. Acronym: could you specify what benefit you see in putting length marks in derived forms/other external links? —ObsequiousNewt (εἴρηκα|πεποίηκα) 14:46, 4 October 2015 (UTC)


 * from as in ;  is the regular Romanisation of the nominative feminine plural form, . The benefit I see in including length marks in Related terms, Derived terms, and Descendants sections is that one is thereby able to see to what degree vowel quantities are preserved or ignored by a term's relations and derivations; that benefit doesn't really apply to Synonyms, Antonyms, and other -onyms sections, however. Apart from that (admittedly not huge) practical benefit, including length marks in those other sections brings greater consistency of presentation; the kind of prescription I've written up is identical in practice to the de facto treatment of Latin terms. — I.S.M.E.T.A. 17:49, 4 October 2015 (UTC)


 * I'm not sure I can think of any examples of a vowel quality being ignored. The only example (but a very important example) that would potentially be relevant is verbs in -ι- or -υ- (or -ρα-, though I can't think of an example of that) that change grade. As far as I can tell the pattern here tends to be: long when suffixed with -μα, -σις (and, judging by e.g. -τηκτ-, almost all productive suffixes); short when in compounds with both -ος (and, by extension, -έω, -ία, -ικός, etc.) and -ης; hidden when in compounds with -τος (but apparently long, judging by e.g. -τηκτος.) So, based on that alone, I find myself in favor of adding derived and related terms to the list. However, I do find myself wanting to stop there, leaving off synonyms/antonyms/see-also/links from external pages. I don't have much of a rationale for wanting to do so, besides that writing length marks is much more labor-intensive than not doing so.
 * Additional comment: the vote seems to me to be poorly written. For one, the first option doesn't oh, I think I see what this is supposed to be saying—in which case "option" was a poor choice of words; "issue" or even "vote" is preferable. Secondly, the second "option" includes language like "where vowel quality is inferable" (which seems to imply that circumflexed letters also need a macron?); it also doesn't take into account links from other pages into Ancient Greek (of which the most important will be translations and etymology). That said, I'll go cast my vote now. —ObsequiousNewt (εἴρηκα|πεποίηκα) 15:51, 6 October 2015 (UTC)


 * D'oh! I tried, but failed, to make the language watertight. Shall I amend the option-2 voting proposal? I would've been fine, BTW, with omitting length marks from non-relational -onym sections; nevertheless, I'm glad to see your vote in support. :-)  — I.S.M.E.T.A. 11:24, 7 October 2015 (UTC)

Macrons in attesting quotations and example sentences
Above, I see ObsequiousNewt say that "quotations should not include length marks". Is there at least consensus that macrons should be absent from attesting quotations and example sentences? --Dan Polansky (talk) 07:13, 3 October 2015 (UTC)


 * I agree with you that macrae: ⟨ ¯ ⟩ and brachiae: ⟨ ˘ ⟩ should be absent from attesting quotations (unless they actually occur in the source text, such as in the "few transcriptions of papyri" mentioned by ObsequiousNewt in, timestamp: 00:58, 29 September 2015). It is in our quotations that we should strive to be as faithful as we can in the reproduction of source texts. I am currently neutral with regard to example sentences. — I.S.M.E.T.A. 13:19, 3 October 2015 (UTC)


 * Out of curiosity, can you clarify what's the business with the terms "macrae" and "brachiae"? Is there a reason why you don't say "macrons" and "breves", to make it easy to understand for as many people as possible? --Dan Polansky (talk) 13:27, 3 October 2015 (UTC)


 * I prefer to use the endonymous words for the Ancient Greek diacritics. represents ;  represents . — I.S.M.E.T.A. 13:58, 3 October 2015 (UTC)


 * It's all Greek to me (pun intended), not English. --Dan Polansky (talk) 14:24, 3 October 2015 (UTC)


 * Well, you have your explanation. — I.S.M.E.T.A. 15:08, 3 October 2015 (UTC)


 * I think that quotations should be as faithful as possible to the source, so it should only include diacritics that the source includes as well, and even spacing should be reflected as they are. A normalised form could also be provided, though currently doesn't have a parameter for it. —CodeCat 23:59, 3 October 2015 (UTC)


 * Agreed. Isn't for example sentences? — I.S.M.E.T.A. 00:07, 4 October 2015 (UTC)


 * Quotations are example sentences too, at least when placed in an entry. —CodeCat 00:14, 4 October 2015 (UTC)


 * I disagree that should be used to present the text of quotations. Its automatic italicisation is undesirable in that context. The use of  obscured the use of italics in this source text, for example. — I.S.M.E.T.A. 00:29, 4 October 2015 (UTC)


 * We terminologically distinguish quotations from example sentences. In Wiktionary parlance, quotations are not example sentences. --Dan Polansky (talk) 05:20, 4 October 2015 (UTC)


 * Keep in mind that there's going to be in most cases multiple "sources": papyri, manuscripts, and translations in scholarly journals; and within the last one you will usually have several scholars who disagree. (Note also that when I have seen macrons [note that I have been using "macra" in the sense of being a plural of "macron"... but I suppose I should use the -s plural to mitigate confusion] they are rarely if ever used to distinguish ᾰ from ᾱ, but rather to note irregularities. A quick skim of Supplementum Lyricum (Diehl, 1917) brings up such markings as ε͡υ, ε̄, ο̄.) —ObsequiousNewt (εἴρηκα|πεποίηκα) 23:04, 4 October 2015 (UTC)


 * Are they not, rather, overlines? — I.S.M.E.T.A. 23:20, 4 October 2015 (UTC)
 * There are more examples of them here. — I.S.M.E.T.A. 23:21, 4 October 2015 (UTC)


 * They are not overlines. Overlines (at least in that context) mark numerals, I believe. —ObsequiousNewt (εἴρηκα|πεποίηκα) 23:27, 4 October 2015 (UTC)


 * In Latin, yes; but Greek uses ⟨ & ⟩ and ⟨ & ⟩, right? — I.S.M.E.T.A. 23:30, 4 October 2015 (UTC)


 * Byzantine numerals used overlines. That said, I'm actually not sure what your example is—those might actually be proper macrons and not overlines; I haven't looked too closely at what he's talking about but it seems to be about the vowels. —ObsequiousNewt (εἴρηκα|πεποίηκα) 23:43, 4 October 2015 (UTC)


 * I'm afraid I can't help much. I do note, however, that there is an overlined nu in line 10 on page 146 and there is an overlined psi (or is it an omega with an iota subscript?) in line 32 of page 147. — I.S.M.E.T.A. 00:17, 5 October 2015 (UTC)

Piling of accents
Here are some examples where accents are piled, i.e. the macron is being added to a letter that already has two accents:
 * ἂ ἃ  ἄ  ἅ  ἆ  ἇ — ἂ̄  ἃ̄  ἄ̄  ἅ̄  ἆ̄  ἇ̄
 * ἒ ἓ  ἔ  ἕ — ἒ̄  ἓ̄  ἔ̄  ἕ̄
 * ἢ ἣ  ἤ  ἥ  ἦ  ἧ — ἢ̄  ἣ̄  ἤ̄  ἥ̄  ἦ̄  ἧ̄
 * ἲ ἳ  ἴ  ἵ  ἶ  ἷ — ἲ̄  ἳ̄  ἴ̄  ἵ̄  ἶ̄  ἷ̄
 * ὂ ὃ  ὄ  ὅ — ὂ̄  ὃ̄  ὄ̄  ὅ̄
 * ὒ ὓ  ὔ  ὕ  ὖ  ὗ — ὒ̄  ὓ̄  ὔ̄  ὕ̄  ὖ̄  ὗ̄
 * ὢ ὣ  ὤ  ὥ  ὦ  ὧ — ὢ̄  ὣ̄  ὤ̄  ὥ̄  ὦ̄  ὧ̄

I consulted Greek_alphabet. From what I can see, while the items before the dash have a unicode point, I need to use the combining diacritic for macron to add the macron. Not sure whether all of the combinations above actually occur. If there are any mistakes, please post a corrected example below.

This shows that adding macrons to Ancient Greek is quite a different thing from adding them to Latin, where there is no piling of diacritics, AFAIK. --Dan Polansky (talk) 06:09, 4 October 2015 (UTC)

Here's one example of added breve: --Dan Polansky (talk) 07:14, 4 October 2015 (UTC)
 * ῠ́ -- which is supposed to be ύ + ̆


 * You display your considerable ignorance of Ancient Greek here, Dan. Epsilon and omicron  are always short, whereas eta  and omega  are always long; therefore, none of those letters will ever have a macra or brachia. Moreover, a perispomene: ⟨ ῀ ⟩ only ever occurs on a long vowel or diphthong, so it will never combine with a macra or brachia, either. The baria: ⟨ ` ⟩ chiefly occurs on the final syllable of non–clause-final oxytones, but also occurs on a few other words like articles; consequently, it will be rare for one to encounter a macra or brachia combined with a baria and, unless I'm very much mistaken, the baria will never occur in combination with a breathing mark (i.e., I don't see what use ⟨ ἂ ⟩, ⟨ ἃ ⟩, etc. can have). Something you haven't mentioned here is the iota subscript: ⟨ ͺ ⟩, but in any case, letters with it are always long, anyway. Besides all that, the technical requirement is that the base letter must first be furnished with the length diacritic, followed by the breathing mark or diaeresis (if it has either of them), and then the pitch diacritic; if this order is not adhered to, the term will not be linked to properly. — I.S.M.E.T.A. 14:33, 4 October 2015 (UTC)


 * I am sure you can do a better job. Please do: show the reader what sort of combinations actually happen, and how they actually appear so the reader can judge for themselves. --Dan Polansky (talk) 14:40, 4 October 2015 (UTC)


 * Why should I undertake your drudgery for you? Work it out yourself; I'm not your underlabourer. — I.S.M.E.T.A. 17:53, 4 October 2015 (UTC)


 * Oh come now, it's not that difficult—except for someone like Dan who, y'know, doesn't know Greek. Here are all of the combinations that should occur: (two diacritics) 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. —ObsequiousNewt (εἴρηκα|πεποίηκα) 22:38, 4 October 2015 (UTC)


 * I was wondering: Can an oxia and a coronis co-occur on a final-syllable monophthong? — I.S.M.E.T.A. 00:20, 5 October 2015 (UTC)


 * Sure; "ἤ τις". Why? —ObsequiousNewt (εἴρηκα|πεποίηκα) 00:40, 5 October 2015 (UTC)


 * I was trying to work out whether any of these: could ever actually occur in a real Ancient Greek word. Can you think of such a situation? — I.S.M.E.T.A. 18:59, 5 October 2015 (UTC)


 * Sure. Observe:
 * Th.1.9: οὐκ νήσων ἐκράτει
 * Hes.Th.304: ἀλλά ἑ ἐδάμασσε βίης Ἡρακληείης.
 * Um. Actually, I couldn't find one with upsilon. Theoretically it could have happened, under any number of circumstances, but it doesn't appear that it ever did.
 * —ObsequiousNewt (εἴρηκα|πεποίηκα) 20:22, 5 October 2015 (UTC)


 * Thank you. Does either ἄν or ἴς occur in non–clause-final position in one or more idiomatic phrases? — I.S.M.E.T.A. 21:12, 5 October 2015 (UTC)


 * ...Yes, in those examples. —ObsequiousNewt (εἴρηκα|πεποίηκα) 22:16, 5 October 2015 (UTC)


 * By "idiomatic phrase" I mean a phrase that could be given an entry here, i.e., that isn't "NISOP" per WT:CFI and vague WT:RFD convention. — I.S.M.E.T.A. 22:23, 5 October 2015 (UTC)


 * Oh, I misread. ἄν almost certainly does; I'm not sure about ἴς per se but there are a bunch of proverbs with . —ObsequiousNewt (εἴρηκα|πεποίηκα) 22:31, 5 October 2015 (UTC)


 * Thanks. OK, that means that ⟨ ᾰ̓̀ ⟩ and ⟨ ῑ̓̀ ⟩ could both occur in our entries. Pretty rarely, though. — I.S.M.E.T.A. 01:07, 6 October 2015 (UTC)