Paul Derouda wrote:Not that I understand very much about this... But aren't the primary endings regular with aorist subjunctive? Couldn't the ε in βλήεται just be an incidental case metrical lengthening, because βλήεται as cited in LSJ appears in a context in the Odyssey that requires -αι to be short? (epic correption before a vowel) If ηε is just metrical lengthening, there's no need for the hypothetical *ξυμβλῆται.
According to my Bornemann-Risch grammar the regular Attic aorist subjunctive is βάληται. So could σύμβληται and συμβάληται just be two parallel developments from an original Indo-European form, the former disappearing first? (like two different developments from vocalic l̥ or something, I don't know this stuff)
About the metrical lengthening: If (ξύμ-)βληται made sense as a subjunctive I'd agree, but the subjunctive marker is missing: *
(h₁)é-gʷl̥h₁-to (indicative) > ἔβλητο and *
gʷl̥h₁-e-toi̯ (subjunctive) > βάλεται/βάληται (should be short-vocalic originally; also, ignore that the original ending was *-
toi̯, not *-
tai̯). Notice that *
l̥h₁ before a vowel (the subjunctive marker!) becomes αλ while before a consonant it becomes λη (with long
ē, not Ionic for
ā).
So you may be right that σύμβληται could have been formed as a replacement for συμβάληται even though the form feels kind of odd.
Another thing I'd like to add: Diphthongs at the end of a word are not really "shortened" before a vowel but the glide element (i̯ or u̯) becomes the onset of the next syllable which makes the syllable before that open and hence short (if it has a short vowel). Example: μοι ἔννεπε should be syllabified as mo.yen.ne.pe.
Paul Derouda wrote:So I'm guessing LSJ is quoting only the "surprising" forms and assumes we know the more common ones. Is this how LSJ works?
Normally LSJ mentions all (or most?) of the attested forms.