This article is within the scope of WikiProject Greece, a collaborative effort to improve the coverage of Greece on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
I don't understand the statement about glottal stops in connection with smooth breathing. It is not possible to start a vowel (from silence) without first closing the glottis, then opening it on the desired vowel.
Even if Greek sentences were or are spoken with no hiatuses, there is a hiatus between sentences. To break the silence on a vowel, you need either a "rough breathing," which results in a voiceless fricative like our "h"; or a "smooth breathing," accomplished by first closing the glottis, then opening it on the vowel. Perhaps the author prefers not to call it a glottal stop. But since he or she specifically mentions glottal stops, he should clarify that he distinguishes glottal closure from a glottal stop.Smokey1066 (talk) 09:32, 27 September 2012 (UTC)
The point Allen makes is that the smooth breathing does not represent a glottal stop. A speaker may choose to insert a glottal stop in some places where there's a smooth breathing, but does not have to. The evidence you give would only work at the beginning of an utterance, not in the middle. In the middle there's no reason to have a glottal stop.
The evidence he gives is that elision can happen when a word ending in a vowel comes before a word beginning in a vowel with a smooth breathing on it. Elision of vowels can only happen if two vowels follow each other, because elision is motivated by ease of pronunciation. If the smooth breathing represented a glottal stop, it would be a consonant. Then there'd be a consonant between the two vowels, and elision would not happen. Therefore the smooth breathing is not a consonant. (Other sound changes involving two vowels, like conversion into an improper diphthong, can also occur in Homer in the same situation as elision.)
For my part, I'm studying Arabic, which definitely has a glottal stop phoneme (for example, in the word Quran, after the r), so I know what having a glottal stop phoneme looks like. There's none in Ancient Greek. — Eru·tuon 14:07, 27 September 2012 (UTC)