Bombichka wrote:if [e]<[ai] merged with another type of e-sound, it merged with the epsilon, forming an open e, and the eta was the closed e.
That's the situation that Randall Buth describes, and which the recording seems to be aiming for.
I think this was so because it is only natural that an a-sound moving towards e retains some of its former wider pronunciation. for example, the e-sound in French coming from an original Latin a (as in lat. amare - fr. aimer) is pronounced more open than the e-sound from Latin e or i, which, in many cases, has even turned to the so-called "e-muet", pronounces like the u in engl. but (cf. lat pensare - fr. peser).
That makes sense -- there's also Vulgar Latin, where ae
merged with short e
to an open sound and oe
merged with long e
(and short i
I guess) to a close sound.
Besides, I recall reading somewhere there are Modern Greek dialects in which there is a distinction between the e's of diphthongic and of monophthongic origin. I still have to confirm this information, but it sounds convincing to me.
If you do confirm it, I'd be interested in that too. The only thing I can think of that's kind of similar is that Pontic Greek has generally preserved Î· as [e], but I don't think that has any bearing here.
As to the second element of the aw- and ew-diphthongs, it is also natural that it should have passed through a bilabial stage before turning into a labiodental consonant. the same semi-vowel has undergone the same development in Spanish. a proof that something like this also happened in Greek are forms such as the dialectal atos instead of autos, where probably a bilabial v before a silent plosive became itself silent and was later "swallowed" by the speakers of the dialect. Modern Greek emorfos < Ancient Greek eumorphos demonstrates a somewhat similar process.
That makes sense as well.