A Hawaiian Long Diphthong Pronunciation Model

[jeidsath]

Allen claims that length differences in the long diphthongs (ᾳ, ῃ, ῳ) versus the short diphthongs (αι, ει, οι) are not sufficient to differentiate them. His proposal (value differences) has struck me as unlikely, because the long diphthongs all evolved together and the short diphthongs all evolved together. Value differences -- I would think -- would have made independent evolution more likely.

I still find it hard to differentiate the long diphthongs in my own spoken Greek.

It turns out that the Hawaiian language has a vowel system similar to ancient Greek, in that it has long and short diphthongs, and a large number of vowel phonemes. Here is an example of aːi versus ai:

ʻāina
ʻaina

While I can't find examples on Forvo, Wikipedia claims that eːi also exists in parallel with ei. There appears to be no oːi, just oi.

[Altair]

Joel, thanks for this info. I have read some about Hawaiian, but had never heard this difference. I can hear it if I pay close attention, but I think my dialect of English (standard American) varies between these two realizations for the same phoneme, especially in open syllables; and so it is not easy for me to hear.

As for oːi versus oi, Japanese does make this distinction, but I think it counts them as two moras versus three as opposed to two and two as in ancient Greek. The long vowel sequence oːi is probably also felt to be two syllables, rather than one; but syllable counting in Japanese is probably not clear.

When I go to speak some languages I think I consistently and naturally use the "short diphthong" /ai/ (e.g., Spanish, German, Arabic) for the equivalent of αι. When I mimic some southern American accents, I consistently use a long diphthong for what I think is comparable to long ᾳ. In my natural English, I distinguish the length of this diphthong between "bite" and "bide," presumably because of the voicing of the syllable coda; but I don't think I can transfer this intuition to Greek.

As my model, I try consistently to use a "short" pronunciation for αι, and then mentally use ā followed by a semivowel "y" for long ᾳ. I have rationalized that since Greek was slowly ridding itself of semivowels over the centuries, this could match what the underlying causative factors actually were. I then can use the idea of a southern American drawl to make the syllable coda lighter and lighter until it disappears into an authentic Koine accent!

[Paul Derouda]

Allen claims that length differences in the long diphthongs (ᾳ, ῃ, ῳ) versus the short diphthongs (αι, ει, οι) are not sufficient to differentiate them. His proposal (value differences) has struck me as unlikely, because the long diphthongs all evolved together and the short diphthongs all evolved together. Value differences -- I would think -- would have made independent evolution more likely.

I still find it hard to differentiate the long diphthongs in my own spoken Greek.

It turns out that the Hawaiian language has a vowel system similar to ancient Greek, in that it has long and short diphthongs, and a large number of vowel phonemes. Here is an example of aːi versus ai:

ʻāina
ʻaina

While I can't find examples on Forvo, Wikipedia claims that eːi also exists in parallel with ei. There appears to be no oːi, just oi.

I’m not really able to tell the difference in length between the a’s in the two ainas. To me it seemed more like the stress was placed differently, on the first syllable with the long a and the last syllable with the short a. At the very least I don’t think vowel length was the only decisive factor in making the difference.

[ariphron]

It's a clear contrast in length, though I had to listen to the "long" version three times to convince myself that there was no significant difference in quality. The short one sounded like a completely normal /ai/ diphthong, while the long one was surprisingly dark (towards /ʌːi/).

No reason not to use this as a model for your Greek, if you can perform the sounds consistently. On the other hand, I don't see a problem with a value/quality difference being present as well, as a secondary feature that makes the length contrast clearer. Isn't that essentially how quality differences between long and short vowels tend to function?

For my own pronunciation, I maintain a consistent prosodic contrast (first vowel fleeting versus held), and a certain difference in quality of the vowels that seems to follow naturally. I suppose I do something of the southern American drawl on the long diphthongs, which helps keep them separate from the short ones, and before consonants the long ones often get slightly pharyngealized. The only argument I can make for the pharyngealization is that it seems like good advice to be resourceful in finding ways to maintain a contrast robustly if you think it's a contrast that is important to maintain.

[Anthony Appleyard]

Dutch has the long-diphthongs "aai" and "ooi". (I taught myself Dutch for two holidays motorcycling around Holland.)