Attic declension and omega

[Cheiromancer]

So I was reading lesson 23 of Ollendorff (p. 86), wherein is given the declension of words like ὁ ταώς (peacock) and τὸ ἀνώγεων (dining hall). Ollendorff says

Gen. Sing. ταώ instead of ταῶ. ἀνώγεων Proparoxytone throughout, ω standing for ο, and being regarded as having but half its usual length.

Now I have been practicing my pronunciation of omicron and omega so that they have the same quality but a different quantity. Omicron as in "obey" and omega as in "own". (If pronounced as the "aw" sound I get it confused with alpha.) Omega standing for omicron really confuses me. If it is has the same quantity and quality as omicron, why didn't they write omicron? But if it is short and distinct from omicron, then it must be qualitatively different. Which suggests that omega was qualitatively different from omicron.

The way I had learned omicron was the sound in "on", while omega was the sound of "own". Correcting this pronunciation promises to make me more sensitive to vowel quantity, which helps in reading verse, but I was hoping that I was moving closer to an Attic pronunciation. But ἀνώγεων makes me wonder about that.

Is there some peculiar explanation for why the Attic Greeks would write omega for the sound of omicron?

[Qimmik]

The Attic declension is the result of a sound change known as quantitative metathesis.

Smyth 34: "ηο, ηα often exchange quantities, be coming εω, εα_." ("_" indicates that the alpha is long.)

http://www.perseus.tufts.edu/hopper/tex ... 99.04.0007

When this occurs, the accent of a proparoxytone word isn't affected--it remains on the antepenult, and doesn't move forward to the penult, notwithstanding that the vowel of the ultima has become long.

But ἀνώγεων is not a standard form. LSJ lists this rare word as ἀνώγαιον or ἀνώγεον, and NT texts seem to record it as ἀνάγαιον.

http://www.perseus.tufts.edu/hopper/tex ... nw%2Fgaion

As for the quantity of the ω of the final syllable, as far as I know it's long, not short. Here's a line from Aeschylus, Suppliant Women 400, in which the meter guarantees that the ω of λεώς is long:

εἴπῃ λεώς, εἴ πού τι μὴ τοῖον τύχοι,

Incidentally, the old Athenian alphabet, in use until 403 BCE (at least officially), represented ο, ω, and spurious ου as O. What we think of as the "Greek alphabet" is the Ionian alphabet, which Athens adopted officially in 403, although it was probably in widespread use in Athens before that date.

http://www.perseus.tufts.edu/hopper/tex ... 99.04.0007

[Cheiromancer]

Thanks, Qimmik. Quantitative metathesis makes perfect sense. I even knew about the phenomenon (and how it left accents unchanged) but I didn't connect it to this section. It is a lot clearer now.