I picked up Ajax after responding to the thread on line 118 and started reading to refresh my recollection.
Line 55: ἔνθ᾽ εἰσπεσὼν ἔκειρε πολύκερων φόνον
Athena is describing the slaughter of the animals. The phrase πολύκερων φόνον is literally “many-horned slaughter”, meaning by enallage “the slaughter of many horned beasts.”
The word πολύκερων seems to be a hapax; at least LSJ only cites this instance. The LSJ lemma reads:
πολῠ-κερως, ωτος, ὁ, ἡ,
But isn’t the accusative πολύκερων a contraction of πολύκερα-ον or the “Attic” declension? If it were a 3rd decl. adj., the masculine accusative would be unmetrical πολυκερωτα, wouldn’t it?
And the proparoxytone accent, printed in both Lloyd-Jones & Wilson’s OCT and Finglass’ editions, suggests that this word behaves like the “Attic” declension, where the accent of the nominative is retained throughout. Chandler sec. 534 lists this as an Attic declension word, and seems to endorse the proparoxtone accentuation.
Am I missing something? Or is this an error in LSJ, uncorrected in the Supplement?
UPDATE: The Cambridge Greek Lexicon gets this right.
If LSJ were right, gen. pl. would be πολυκερωτων. And gen. pl. wouldn’t make sense: the slaughter of many-hornèd beasts, i. e., the slaughter of beasts with many horns. But each beast has at most just two horns. So it must be acc. sing. agreeing with φονον: a many-hornèd slaughter, meaning a slaughter of many beasts with horns, by a well-known rhetorical figure common in tragedy, enallage, more or less.
It’s surprising that the error escaped correction in the LSJ Supplement, surprising that it even occurred in the first place—especially since just a few lines later in the play (64) we have εὔκερων ἄγραν.
LSJ certainly won’t have been taking it as genitive plural!
The tiny LSJ entry is buried away ina long string of πολυ- compounds, so it easily escaped anyone’s notice. Most readers would not bother to look up the transparent compound, anyway.
I saw the “of” and was hazarding a guess as to what the LSJ was doing. Anyway, the Supplement apparently fixed exactly the same problem with ταυρόκερως. χρυσόκερως is listed with -ωτος first, though it has the form only once in Josephus. μονόκερως also has an -ωτος form in the LXX, but the meaning is different enough that it gets an separate “II” header in the entry.
The online versions of the LSJ get the accents wrong all over for these. Not a problem with the print version.