Thank you all so far for your votes and some of you for your opinions, I'm hoping more of them (votes mainly) will come
In my original post I don't distinguish between "rising tone" and "higher tone", but I guess it was concluded...
Acute: short: high tone; long: low-high combination(=acute on the latter mora. sounds like a raising tone)
Grave: always low/mid tone (i.e. no heightening, flat)
Circumflex: high-low combination(=acute on the former mora, sounds like a falling tone)
Good, that seems to be identical with what I have deduced later on from my own practice!
Under certain circumstances I would pronounce the acutes and the circumflexes on long vowels as pure raising/falling tones though. e.g. ὦ
The problem for me here would be probably, that in order to produce the rising and falling tone in one syllable (2 moras) it would require even more time (marking audibly the transition would consume more time). Even 'jumping to the higher tone and getting back' already seems to make a normally long vowel even longer... But interesting!
PS: I've read somewhere that the grave diacritic was once written on ALL "un-toned" syllables, so I think that that the grave accent is a tone unchanged in pitch (in other words, ancient Greeks simply did not pronounce the acute accent of the last syllable of words being immediately followed. i.e. there was a tone sandhi) is correct.
Yeah, that's basically the best argument there is (from what I've read to this point) for some practical advice for the grave accent (I expressed it by the word "neutral" in my own post)...
I also thought that it wouldn't be a bad experiment to do a research if the grave accent evolved into the same kind of dynamic accent in "grave accent phrases" as it evolved in "non grave accent phrases"... but it would have to be probably a phrase constantly in use through the language evolution (i.e. if we can deduce something about the grave accent diachronically using the today's Greek - but probably the result/the conclusions would be fallacious).
I try to use the tonal accent as far as I can:
- acute: for a short vowel, high; for a long vowel, rising.
- grave: treat it as if the original acute is ignored, i.e., just flat. I see some authors simply drop it.(http://community.middlebury.edu/~harris/Archilochus.pdf
- circumflex: high and the falling, (in my mind it's rise-and-fall but that's not easy to do in practice, unless you do the long vowel unnaturally long)
My native language is Czech (western Slavic language, IE) and we have
phonemically long and short vowel pairs (something quite rare in Europe and quite 'classical-languages connecting'), so I naturally happen to distinguish long and short vowels (even in a rapid speech: we never get this wrong) and I also naturally produce them in the desired length and I must absolutely agree
with your observation about the vowel length of the circumflex syllables! It indeed seems to me (from my tonal pronunciation realizations I have done to this point), that if I want to make the circumflex "real" (to apply properly the two tonal changes on each mora) I have to make the long vowel at least by 1/3 (or 1/2) longer that what is for me a "normal natural length of a long vowel". And I'm glad that we, though disconnected from each other normally and having different lingual backgrounds, arrive independently at the same result
I also try to pronounce iota subscript as an i-glide for memory's sake, (also justified by its use before 5th century BCE as explained in Vox Graeca.)
Yes, that would be an interesting poll again: how do you normally read subscripted iota. But for me personally (the little bits I have read) there seems to be a clear recommendation (for the reconstructed pronunciation users) for pronouncing it (if that is what you say... maybe I've got lost in some terminology you have used)
I have done so far only a "quick" reading of Vox Graeca (mainly of the parts that got me interested the most).
(So right now I pronounce all the subscripted vowels as long diphtongs)