Several times in Homer a seemingly short syllable scans long before νέφος (e.g.,. Od. 9.145), and I don’t know why.
οὐρανόθεν προύφαινε, κατείχετο δὲ νεφέεσσιν.
Was there originally another letter before the nu, a sigma or a gamma? Digamma does not occur before nu.
Chantraine doesn’t give any suggestion and the cognates (e.g., nebula) begin with a /n/. Any thoughts?
As someone who has followed the ζ=dz arguments more closely than is good for him, the crazy idea of δνεφος comes to mind. Googling brings up articles discussing it. I’m sure there are more traditional explanations for it though.
I see that δνοφος (gloom in LSJ, eventually γνοφος and storm-cloud for Aristotle) does not occur in Homer, but ἰοδνεφὲς does and δνοφερὸν.
Interesting line:
σκίδναθ’ ὑπὸ νεφέων, τανύοντο δὲ μώνυχες ἵπποι
Homer always uses variants like this for σκεδαννυμι (except aorist).
D.B. Monro A Grammar of the Homeric Dialect:
https://dcc.dickinson.edu/grammar/monro/lengthening-ρ-λ-μ-ν-σ-δ
or the scanned version from the Internet Archive:
https://archive.org/details/grammarofhomeric00monruoft/page/344/mode/2up
Here is a link to a post where consonant doubling is discussed (along with other questions):
http://discourse.textkit.com/t/odyssey-reading-group-no-such-thing-as-a-stupid-question/16848/1
Unfortunately, if you read through, he does not give any suggestions for the origin of the νεφος lengthening. (It’s not σνεφος, like some of the other ν-words.)
A Latin cognate like nebula could have a last common ancestor with νέφος that was after the loss of a consonant, but there’s also Sanskrit nabha, which would have to have split off much earlier. Beekes assigns a PIE root that starts with n.
There’s an incredibly detailed commentary by Anthon on the first six books of the Iliad, and it includes a metrical index where he tries to explain every doubtful line. Unfortunately it seems to cut off after book 3: https://archive.org/details/firstsixbooksofh01home/page/446/mode/2up I don’t find any instances of νέφος until 4.274.
If you want to get really crazy, you could see if you could substitute some form of κνέφας, darkness, and preserve the sense. In a cloud → in darkness, etc. But of course that way would lie madness…
Joel’s δνοφος is ingenuous but νεφος doesn’t seem to come from the same original root, according to LfgrE.
In epic, word initial continuant consonants (λ μ ν ρ σ) can be lengthened and so lengthen the preceding open syllable, without having to resort to any kind of “etymological” explanation. But according to West’s Introdution to Greek Metre this even applied originally to ϝ, which explains μέγα ἰάχουσα = me-ga(w)-(w)i-(w)a-khoo-sa.
I am indebted to Catherine Roth for this reference.
Meillet & Vendryes, Traité de grammaire comparée des langues classiques, referring to words beginning with sn-:
Par analogie, le poète étend cette faculté à des mots qui originellement commençaient par ν simple; ainsi ἅμα δὲ νέφος Δ 274 (cf. O 625, υ 114) compte pour … bien que le mot νέφος n’ait jamais perdu de sifflante à l’initiale (skr. nábhah <>).
Analogy (which of course destroys linguistic regularity) probably does make the best sense as an explanation for the νέφος anomaly, pace Joel and the clever δν- idea. Thanks for the references.
I don’t think that δν- deserves to be called anything but a spitball sort of idea. But the real difficulty with accepting analogy as the explanation is that this initial lengthening is nearly regular for νέφος in Homer. Out of 33 appearances in the Iliad, only these three force a short initial ν (by my count):
13.523:
ἀλλ’ ὅ γ’ ἄρ’ ἄκρῳ Ὀλύμπῳ ὑπὸ χρυσέοισι νέφεσσιν
17.243:
καὶ σῇ, ἐπεὶ πολέμοιο νέφος περὶ πάντα καλύπτει
17.372:
ἠελίου ὀξεῖα, νέφος δ’ οὐ φαίνετο πάσης
Further, I’ll expand my constellation of strange words, including bcrowell’s suggestion, as well as a few others I trudged up from LSJ-peeking. All of these have that doubled initial constanant and an apparently atmospheric connection.
δνόφος (and derivatives) - darkness
γνόφος (in Aristotle’s time) - from δνόφος
–ἰοδνεφής (presupposes *δνεφος, says Beekes !!)
ζόφος - darkness (δσόφος stage possible?)
κνέφας - darkness
ψέφας - darkness
δνόψ - “χιτῶνος εἶδος, βάθος, Hsch.”
Δνεύς - (dialect for Ζεύς, some kind of sky-god I hear)
ζέφυρος - west wind (δσέφυρος stage?)
πνέω - blow
Separately: νιφάς - snowflake, regularly lengthened in Homer, but “sn-”, related to our “snow”, etc.
Beekes calls most of these pre-Greek (πνέω onomatopoeic). Given all of the cognates, maybe νέφος really is a different root with an initial n- sound. But maybe νέφος forced some other word with a very similar meaning out of the text/poetic tradition (ie., that postulated δνεφος, meaning something like storm-cloud?).
Well, I agree analogy is a bit weak as explanation here, because I think in general initial continuants are often lengthened after a word-ending open syllable. It’s possible that it’s more common with ν than with other continuants (though I really don’t know) and it’s possible that that’s partly due to analogy, as claimed by that citation on French, but the problem is that it doesn’t explain cases like μέγα ἰάχουσα.