Talk:ἀήρ

@Erutuon A long ā would not reflect the initial laryngeal of the root. Initial laryngeals always vocalise to short vowels. —CodeCat 23:46, 25 March 2016 (UTC)

For comparison: ἀέξω (aéxō) is listed as derived from *h₂weg-, which has the same initial consonants. —CodeCat 23:50, 25 March 2016 (UTC)

Quite right. But in ᾱ̓ήρ (āḗr) the long vowel comes from the Proto-Hellenic cluster *hw, which turns to a geminate semivowel, written υ (u), in Aeolic, but lengthens the preceding vowel by compensatory lengthening in Attic, Ionic, and Doric. It's the same process that caused *ehmi to change to Aeolic ἔμμι (émmi), but Attic-Ionic εἰμί (eimí) and Doric ἠμί (ēmí). It applied to all clusters with PG *h from PIE *s and a sonorant. This is discussed in Kiparsky's Sonorant Clusters in Greek.

I suppose this entry and others should be updated to explain this... though it would be a lot of work. — Eru·tuon 23:54, 25 March 2016 (UTC)

But where is this h from s? The root is simply aw- not ahw-. By the way I've created a discussion at WT:ES about this. —CodeCat 23:58, 25 March 2016 (UTC)

Oh..... I was confusing this word with ἠώς (ēṓs). So the sonorant cluster thing doesn't apply here. But still, Cunliffe says the word has long alpha. And if I do a little search of the text of the Iliad and Odyssey, the alpha is certainly always long. And Latin borrowed the word with long a. Not sure why, since it's not from **ahwēr. As you say, it should be short. — Eru·tuon 00:27, 26 March 2016 (UTC)