Mora based pronunciation for ancient Greek

[jeidsath]

Mora based pronunciation Iliad opening

So, I think that I've found out the answer to my Greek poetry question. Greek in the classical period was a mora-based language (like Japanese). Allen mentions this in his accent section of Vox Graeca, but it doesn't inform the rest of his work.

I've been working on mora-izing all of my long vowels all the past week, and while I still can't speak quickly, you can hear the results above. (If anyone wants to do the same, I highly recommend the exercises from this video: https://www.youtube.com/watch?v=jSIwo5v5vnw -- transferred to Greek, of course.)

My results so far: 1) The accent rules are emergent (even if the accent isn't). In order to pronounce all of the long vowels as exactly doubled short vowels, I seem to be doing a subvocal pitch variation (ie., like the circumflex, but not pronounced aloud.) In a succession of rolling vowels, there are a only a few places where the up-pitch (or low deep pitch) can be emphasized. To see what I'm talking about, try pronouncing the following succession of short α together as fast as you can (as one word): ααααα. Practice until the stops are gone.

2) The sound of quantity-based poetry is emergent for the mora-ized vowels. You don't have to emphasize any of the beats. The long vowels are *not* prolonged short vowels. They are doubled short vowels.

3) The standard advice for how to pronounce diphthongs (and long vowels) is terrible. The quality is the least important part. They are all doubled vowels. Long-α, etc., are similar to the diphthongs, but without a variation of quality. Short diphthongs are possible (word-ending αι and οι in Attic), and I pronounce my iota subscripts as a normal doubled vowel, the first half is the initial vowel, and the second half is the short version of the diphthong. The length is exactly right, and the quality difference is perfectly intelligible to the ear.

1b) The pitch accent seems to be just an exaggeration of the natural vocal chord movements, if you're doing it right. For an idea of the variation possible for a native speaker, listen to this (especially the snake sounds): https://www.youtube.com/watch?v=7pakfqh-XUI

I imagine similar things are true for Sanskrit (certainly) and classical Latin (possibly). The svarita makes complete sense in light of the above.

Also, if you are using Japanese as a model, it is unfortunate that they have very few double-vowel sounds, since nearly every mora in the language is a consonant-vowel pair. But you can look up words like Osaka or Tokyo on forvo.com for good examples of long vowels. Also compare the words for grandfather (grandmother) to uncle (aunt). The only differences are the prolonged vowel.

[jeidsath]

To try to outline this a bit better, another model would be Estonian, where vowel length is reinforced by a pitch contour. Listen to their President's name (Thomas) for an idea of how long I think that double-vowels were in ancient Greek.

My Russian/Georgian friend, when I asked him whether Estonian "sounds slow," told me "the joke about Estonians is that the new erotic television show is going to be 'an hour there, an hour back.'" It makes sense to him if not to me.

EDIT:

Estonian poem

Notice the pitch and doubled vowels.

Nutikas lutikas
liugu lõi jääl,
nahk sai tal märjaks ja
ära läks hääl.
Järgmisel hommikul
oligi tumm.
Kõht aga paistetas
suureks kui trumm.
Nutikas lutikas
voodisse jäi.
Üles ei tõuse,
söömas ei käi.
Tulid poisid,
tegid pai.
Nutikas lutikas
terveks sai.

And some prose from the UCLA language archive:
http://archive.phonetics.ucla.edu/Langu ... 69_01.html

[Paul Derouda]

Estonian actually distinguishes three degrees of vowel length and three degrees of consonant length! (in accented syllables only, I think) So the model for Ancient Greek you have chosen is not exactly easy... But perhaps indeed pitch in overlong syllables is a sort model for the circumflex accent, or what do you think of this random abstract I found by Googling a bit: http://link.springer.com/chapter/10.100 ... -0137-3_10.

Estonian is pretty close to Finnish. When you hear someone speaking Estonian, it's just like Finnish, except that it sounds very funny and you don't exactly know what they're talking about, you understand two words in three but not enough to get the meaning. And when Estonians speak Finnish, it still sounds exactly the same (funny) except that now you understand! (Estonians who lived on the North coast and were old enough to watch TV during the Soviet era generally know Finnish well, because many of them preferred the capitalist lies broadcast by the Finnish television to the local TV run under the paternal surveillance of the Communist Party. Nowadays there are many, many Estonians working in Finland, because the wages are three times superior, the languages similar and the distances between the countries minimal.)

Anyway, I guess I just learnt one reason why Estonian is so funny: the way the pitch contour works. I don't think Finnish has anything like this, or at least I'm not aware of it. For vowel length at least I would say that Finnish is a better model because it's simpler, though. Three grades of vowel lenght, honestly!

[Paul Derouda]

Re:morae it seems to me you're making things more complicated than they are. I think Finnish is described as a mora-based language by some (though I don't have any notion of Japanese, so I can't compare), and as far as I know long vowels are just like short vowels except longer (and not necessarily exactly two times as long). Analysing them as two consecutive shorts might be useful from a theoretical point of view, and certainly diachronically long vowels often are a result of two short vowels becoming one when the intervening consonant has dropped out (e.g. talohon -> taloon "into the house"). The difference with Estonian is (I suppose, not sure as I just learnt this) is that Estonian plays around with the pitch contour in a way Finnish doesn't.

This might be useful, I have already posted this on Textkit earlier. I demonstrate the minimal pairs a/aa and k/kk.

https://www.youtube.com/watch?v=MaUYWRe4OB0

The words are: taakka -- takka -- taakkaa -- takkaa -- takaa

(Meanings: burden -- hearth -- burden (partitive) -- hearth (partitive) -- from behind.)

This Wikipedia article might be useful too: http://en.wikipedia.org/wiki/Chroneme. (The article actually maintains Finnish has three distinctive vowel lengths, which is something I've never heard before. Before now, I've seen the examples given explained in another way, never by resorting to that explanation. Hmm.)

[jeidsath]

I've been reading up on Finnish, actually. Antti Iivonen's article in Intonation Systems, a Survey of Twenty Languages, as well as listening to your audio from the audio thread (and listening to Finnish on Forvo.com and the UCLA site). And thank you for the YouTube link, that's very useful!

Here is what Iivonen has to say about pitch variation and its relationship to vowel length in Finnish:

It has been difficult to find stable microprosodic pitch differences combined with the single/double contrast of vowels. The only difference observed by Aulanko (1985) was a greater F0 movement in the double (long) vowels. Vihanta (1988) concludes that the systematic F0 movements observed in connection with the single/double contrasts in vowels and consonants might depend on word structure or sentence intonation, but they might also have some function as an acoustic cue for the quantity opposition.

There are a few numbers given above about the variation in Finnish vowel length (which seems bigger than the average ratios that I could find for Estonian). It seems to depend on word position and consonant pairings.

For Greek, everything seems to boil down to pronouncing all of my long vowels with the same length that I give to the high -> low tone movement on a circumflex. So listening to your Odyssey on YouTube, I'm trying to give most of my long vowels and diphthongs the same length that you are giving Μοῦσα, βοῦς, κτλ. I'm curious though if you notice a difference between the length of those vowels and others (especially your initial diphthongs, which strike my ear as slightly shorter than the other longs).

[Paul Derouda]

Don't listen to that Odyssey recording too much... For one thing, ου in βους is much too long, as I'm struggling with the rhythm at that point. Whether it's right or wrong, I pronounced long vowels right before the caesura longer. With η in ψυχην it worked better in my opinion.