User talk:Gilgamesh~enwiktionary/Greek IPA tokens

I have some remarks about your table. Firstly, it's too long. I don't know if there's a reason to distinguish between ᾰυ and ᾱυ and so on and so on. Two rows, in my opinion, should be enough.
"β pronounced as f Before voiceless consonants": Can you give an example? I cannot. The same applies for "γ as x before voiceless consonants", "γ as ç before voiceless consonants that have a Contemporary palatal pronunciation". If you have something in mind that I can't think of, please help me.
Finally (for the moment) there is the "palatal before palatal" issue. I insist that n and l don't become palatal before another palatal and I asked you to read the relevant chapters of P. Mackridge's book (if you can find it of course). If you have read another book with something different, please give me the title to see if I can find it. --Flyax 19:35, 17 September 2007 (UTC)

Well, it's not a guide. It's for templating. So that you do something like {{grc-ipa-code|e|leu$|th|e|r|o'|s}} and automatically generates multiple IPA guides. For instance, for Contemporary, it goes about, one by one, embedding the templates {{grc-ipa-con-e}}, {{grc-ipa-con-l}}, {{grc-ipa-con-eu$}} by way of embedding the template arguments one at a time by way of {{grc-ipa-con-{{{1}}}}}, {{grc-ipa-con-{{{2}}}}}, {{grc-ipa-con-{{{3}}}}}, etc. So it's not an auxillary IPA code to appear as itself in pages like with SAMPA. It's only a template-level code that's processed by an orthogonal template network and never actually seen embedded in a page. - Gilgamesh 20:03, 17 September 2007 (UTC)

Oh, I misunderstood some things. Well, ᾰυ and ᾱυ are short and long diphthongs, contrasting with ευ and ηυ. I had understood that these were like the short and long diphthong distinctions for αι and ᾳ, ει and ῃ, οι and ῳ, etc. I had also thought of ου and ωυ, but I remember now that ωυ does not survive as a single diphthong in Classical Greek and becomes ωϋ. So are you suggesting that there is no short/long distinction for Attic diphthongs that end with υ, so that Classical αυ is simply [aʊ] and ηυ is simply [ɛʊ] in all situations? Then why does it become [iv] in Contemporary? As for voicing and devoicing of consonants, I had also understood that this a natural Greek sandhi law. So this is not true? What about ἔκζεμα (ékzema)? That is not Ancient [éɡzdema] and Contemporary [ˈe̞ɡze̞ma]? And what about Greek τζ being [dz]? - Gilgamesh 20:15, 17 September 2007 (UTC)

(i) About ᾰυ and ᾱυ: if you just wanted to create a code, then I have nothing else to say. I don't know if there were or not an ᾱυ in Ancient Greek. I just wanted to say that it is not necessary to distinguish them in the entries, unless of course there are reliable sources.
(ii) You are absolutely right about ἔκζεμα (ékzema); but this was not my point. I mentioned three specific rows in your table, and I'd like to keep the discussion only to these rows-for the moment. So, do you have any examples so as to have a stable basis to proceed? Think of άγχος, αγχίνοια; it's ŋ, neither x or ç. --Flyax

Well, alright. I cited ἔκζεμα (ékzema) because it's an example of a consonant changing its voice to conform with the consonant that comes directly after it. Voiceless before voiceless, and voiced before voiced. As for voiced becoming voiceless, αυ is an allophone of αβ, right? So if αυτό is [afto], then (hypothetical—not real word) αβτό wouldn't also be [afto] but instead [avto] with an enunciatedly voiced [v]? If I come up with a stable example, I'll let you know. - Gilgamesh 21:33, 17 September 2007 (UTC)

Oh, and this concept is reinforced (whether or not it may actually be misleading—I never thought it was misleading) by certain standard romanization systems for Greek. The United Nations (UN) system spells αυ as "af" before voiceless consonants, but the U.S. Board of Geographic Names (BGN) system always spells it "av" regardless of context. (BGN isn't a totally shallow system though—γ is romanized "gh" or "y" depending on what vowels follow, and ντ is "d" or "nd" depending on context, μπ is "b" or "mb" depending on context, γκ is "g" or "ng" depending on context, etc.) The first capital of Modern Greece, Ναύπλιο is "Navplio" by the BGN system. - Gilgamesh 00:11, 18 September 2007 (UTC)

We discuss about the pronunciation of the letter "β", not the romanisation of "αυ". You have a number of rows in your table for "αυ", so we don't need to say anything more about that. My question is very simple: Is there a word containing the letter "β" before a voiceless consonant? Because if, as I tend to believe, there is not, then this specific row should be removed. --Flyax 18:34, 18 September 2007 (UTC)

Well, that sounds wise. I suppose what seemed like a sensible unspoken conclusion to me may have bordered on original research. - Gilgamesh 18:40, 18 September 2007 (UTC)

Right now I can think of one example, and it's purely etymological, not phonological. Μυζηθράς to Μυστράς, where in contraction, ζ before a voiceless consonant becomes like σ. That's probably a rather weak example though. I'm still looking out for anything else. - Gilgamesh 19:29, 18 September 2007 (UTC)

Tell me this, though. How would you instinctively pronounce Νάζκα (Greek for Nazca)? [nazka] or [naska]? - Gilgamesh 19:31, 18 September 2007 (UTC)

I'd pronounce it instinctively with a z. Could you write for me its pronunciation in English or/and Spanish?

English [næzkə], Spanish [naθka] or [naska] (depending on dialect). - Gilgamesh 21:20, 18 September 2007 (UTC)

It turns out that we can look for letter combinations in http://www.komvos.edu.gr/dictionaries/dictonline/DictOnLineTri.htm by specifying e.g. %βκ% (this is the other copy of the dictionary of the Triantafyllides Foundation). A run through of the combinations β + θ/κ/ξ/π/σ/τ/φ/χ/ψ turned up one word, αβτζής which is probably not indicative because τζ -> dz. (pronunciation listed: [avdzís]). Without concrete examples of words in the language, I also think it is a good idea to leave things like this out of the table. ArielGlenn 01:20, 19 September 2007 (UTC)

I've already stricken them out, which should be clearly visible in the table. I'd have commented them out instead, but that's not that easy in MediaWiki with template-based tables. I didn't want to totally delete them until we were most absolutely certain about this. I mean, there still may be biblical names (in the Septuagint or New Testament) where /[βδζ][θκξπσςτφχψ]|γ[θπσςτφψ]/i could exist, and it would be nice additionally to know how they are pronounced in Greek. - Gilgamesh 03:14, 19 September 2007 (UTC)

I was able to get both of these texts from el.wikisource.org (expoprt by category) and grep through them for the above combinations surprisingly quickly! The results are four names: Μαβσάμ, Γεδσών, Γεδσούρ, Βηθσεεδτά. So now that we have them... what do we do with them again? ArielGlenn 09:29, 19 September 2007 (UTC)

In ancient Greek words a βσ combination would have been merged into ψ (see Ἄραβ-ς > Ἄραψ). There is also the rule in Ancient Greek "dentals are expelled before σ" (ὄρνιθ-ς > ὄρνις, τάπητ-σ> τάπης, ἀσπίδ-ς> ἀσπίς). It's apparent that these rules don't apply 100% to words from other languages, mainly to proper names. The most likely explanation is that authors intended to indicate the original pronunciation in Hebrew. So, if I was obliged to utter these strange words, I would hesitate for a while and then I'd simply pronounce what is written: /vs/, /δs/, /δt/. Gilgamesh, do you happen to know these words in Hebrew? How are they pronounced? ArielGlenn, thank you for your research.--Flyax 11:25, 19 September 2007 (UTC)

Oh, you know I'm good with Hebrew? :3 Well, Hebrew was first specified in its own language with concrete vowels in the 9th century in the Masoretes in Tiberias, so some of the Septuagint names use more archaic vowels. But using Tiberian vowels, I do recognize Μαβσάμ as being מִבְשָׁם which is transcribed academically as Miḇśām, which is pronounced [miβɬɒːm] or [miβsɒːm] (depending on whether or not שׁ lost its traditional [ɬ] pronunciation and acquired the [s] used in all major modern forms of Hebrew). In Standard Sephardic Hebrew of Israel, it is transcribed Mivsam, and in both Standard Sephardic pronunciation and in Israeli simplified pronunciation it is pronounced [mivˈsam]. Traditional Yiddish-speaking Ashkenazi Jews would pronounce it [ˈmɪfsəm]. But if I had to take a gander at reconstructing the Septuagint-era pronunciation, I would educatedly guess [məβɬɒːm]. These are from an older theoretical form *[mabɬɒːm]. Does that explain anything? - Gilgamesh 13:04, 19 September 2007 (UTC)

The only conclusion I can reach to is my previous one; that the authors of the Greek text of the Bible wanted to indicate the Hebrew pronunciation of the word. So, the contemporary pronunciation of βσ is just /vs/. I don't think we should bother anymore about that. --Flyax 17:29, 21 September 2007 (UTC)

BTW, I removed lj and nj from the table, and I'll get around to fixing it in the Classical Contemporary IPA entries. As for books...sorry, no. I don't really have access to obscure books or libraries that would have them. I honestly can't even get out of the house much due to my handicap. I might as well live in Antarctica for the non-electronic academic reading material quickly available to me. Wiki editing is a hobby while I'm staying out of direct sunlight (allergic) during the daytime and everyone is otherwise asleep at night. But I've combed most books I could reach on hand since I was a teenager. It all gives me something to do. - Gilgamesh 20:24, 17 September 2007 (UTC)

I combed through all the entries linked to by {{grc-ipa}} and replaced [ʎʝ ʎc ʎç] with [lʝ lc lç]. It was easy and quick—there were only a dozen or so entries that had it. - Gilgamesh 16:00, 19 September 2007 (UTC)

I am trying to get a hold of a copy of the Macridge book (in English) through a local library so we can at least all have a look at the section on phonology. Perhaps this will help resolve some of these issues to everyone's satisfaction. It may take a little while though. Do you folks think this is worthwhile? ArielGlenn 01:26, 19 September 2007 (UTC)

Most definitely. Please. - Gilgamesh 03:09, 19 September 2007 (UTC)

I am sorry to hear about your health. I suppose that you could find some good books in Amazon.com or somewhere else. I am sending you my best on-line-wishes.--Flyax 21:07, 17 September 2007 (UTC)

Well, thank you. But it's not like I'm in an illness... I was born with the handicap. I work with it on a daily basis. Yeah, it makes some things difficult and impractical, but it's okay. Really. ^_^ Thanks for your concern. - Gilgamesh 21:33, 17 September 2007 (UTC)

Testing. View source to see. (deprecated template usage)

• (5th BC Attic) IPA^(key): /ele͜ʊtʰerós/

• (1st BC Egyptian) IPA^(key): /ɛlɛʍtʰɛrós/

• (4th AD Koine) IPA^(key): /eleɸθerós/

• (10th AD Byzantine) IPA^(key): /elefθerós/

• (15th AD Constantinopolitan) IPA^(key): /elefθeɾós/

- Gilgamesh 07:02, 21 September 2007 (UTC)

ngj / a

There is a mistake in row: "ngj, γγ" is not pronounced as ɲɟ "before front vowels in Contemporary pronunciation". We have already discussed it a lot and the pronunciation is ŋɟ. Since I have P. Mackridge's book only in its Greek translation, I'll try to re-translate the following lines (p.69-70, chapter 1.4.2.4)
"Πριν από τα οδοντικά εμφανίζεται το [n], πριν από τα διχειλικά το [m], πριν από τα υπερωικά και ουρανικά το [ŋ] και πριν από τα χειλοδοντικά το [ɱ]. ... ανάγκη /anángi/ [a'naŋɟi]"
"Before dental consonants we have [n], before bilabial [m], before velar and palatal [ŋ] and before labiodental [ɱ].... ανάγκη /anángi/ [a'naŋɟi]"
I know, it's a lousy thing to try to translate in English the Greek translation of an English text :) --Flyax 18:27, 21 September 2007 (UTC)

Sources are the key. ^_^ Fortunately, I anticipated potential future misunderstandings or disagreements on pronunciation, and so I designed the token templates to have a centralized hierarchy of dependencies. Fixing these things will often depend on changing only one template. And this is the template in question: {{grc-ipatok-ngj'}}. Easy. - Gilgamesh 20:04, 21 September 2007 (UTC)

...and it's done. :3 - Gilgamesh 20:05, 21 September 2007 (UTC)

OK. Now, may I ask you why did you change "a" to "ɐ"? I am not sure I know the difference between them. There is a vowel chart here and it seems that we need another symbol, now missing, for modern Greek a. something between /a/ and /α/, but not higher (ɐ)--Flyax 20:32, 21 September 2007 (UTC)

Well, as I understand it, a lower central articulation than [ɐ] is deemed impossible, or at least impossible to practically distinguish from [ɐ]. That's precisely why a lower symbol is missing from the table. It has something to do with oral biology, and...it's complicated, isn't it. XD I don't even like the appearance [ɐ]—it's ugly. But IPA articulations aren't always about what's pretty, are they? :3 (BTW, the IPA back vowel symbol is [ɑ]—it took me about half a minute to realize what you meant. Certain Greek letters were modified in appearance to appear more latinate (or more strictly uniform between typefaces than Greek typeface variations typically permit) before they were adopted as official IPA symbols. Also compare φ with [ɸ], and γ with [ɣ].) - Gilgamesh 21:02, 21 September 2007 (UTC)

Greek α is always described as open central. Could you please take a look here and tell me your opinion? Mainly about that: Since no language distinguishes front from central open vowels, a separate symbol is not considered necessary. If required, the difference may be specified with the central diacritic, [ä]

As far I am concerned, I would prefer to continue using /a/, as the most simple solution. --Flyax 21:15, 21 September 2007 (UTC)

I edited {{grc-ipatok-a}} to use [ä]. The other templates depend on that one. ^_^ - Gilgamesh 21:34, 21 September 2007 (UTC)

Actually, it occurred to me that [a e o] with all those diacritics is a bit visually cluttering, and may actually hinder clear understanding. I do recall that, while as you said [a] suffices for central open, [e o] suffice for pure-mid vowels. I also changed Byzantine [i(ʷ)] to [y], since it was largely [y] before the 10th century, and afterwards it was [i] (or [u] in Aegina, Athens, Cyme, Mani and Megara—Athens apparently had some revived medieval existence before it vanished again and was absorbed into Arvanite culture, and modern Athenian Greek seems to have come from other Greek-speaking Greeks who moved in after the city was refounded in the 19th century). I will examine the IPA and ponder other ways I have made the symbols too fussily detailed. - Gilgamesh 22:14, 21 September 2007 (UTC)

Please see Category:Ancient Greek IPA tokens. It's ready. ^_^ - Gilgamesh 21:45, 22 September 2007 (UTC)