8

It's been 4,000 years since I studied ancient Greek, and the only words I can remember that once contained digammas started with the letter: οἶνος, οἶδα, and so on.

Were there word-internal digammas of whose existence we have evidence? If so, what are some examples? (I'm particularly but not solely interested in digammas that might have immediately followed a gamma or a kappa.)

0

3 Answers 3

5

Like all centum languages, Greek lost the contrast between sequences of a velar + *w and labiovelars, so you're not going to find instances of a velar followed by a digamma: in all Greek dialects of the alphabetic era, those would have become π/β/φ, τ/δ/θ, or κ/γ/χ depending on the dialect and context.

The reason you only remember words beginning with w is because those are just the most prominent on account of their obvious effects on metre in (primarily) Homer. TKR gave some examples of words with intervocalic digammas, but one famous word with a non-intervocalic word-internal w that we can be quite certain of is Attic κόρη, Ionic κούρη < Proto-Greek *kórwā, as well as the male equivalent Attic κόρος, Ionic κούρος < *kórwos (the difference between the dialects here is entirely due to the different way that w was lost), which are attested in Mycenaean as 𐀒𐀷 ko-wa and 𐀒𐀺 ko-wo respectively. (I'm not aware of any direct attestation of the digamma there in alphabetic Greek, though.)

8

A famous one comes right from the beginning of the Iliad, where the digamma in μῆνιν ἄϝειδε θεὰ prevented the vowels from contracting together.

For examples that occurred after consonants, Cairnarvon mentions κούρη < κόρϝα; another famous one is ξείνος < ξένϝος. The vowel was lengthened by the digamma's disappearance.

7

There are quite a few, e.g. νέϝος, κλέϝος, ἔλαιϝον, λέϝων, Διϝός, ἑσταϝώς and other perfect participles in -ϝώς, χαρίϝεις and other such adjectives in -ϝεις, ἡϝώς, αἰϝεί.

I don't know of many examples where *w followed a velar, but where it did, the outcome was similar to that of a labiovelar; the best-known example is ἵππος where -ππ- < *-ḱw-, and there is also πᾶμα (possibly *ḱw- though this is debated) and maybe others.

7
  • 1
    The outcome of *kw is exactly the same as that of *kʷ in all instances; people who see ἵππος as a regular reflex of *h₁éḱwos blame the palatovelar.
    – Cairnarvon
    May 8, 2023 at 21:21
  • 2
    @Cairnarvon I don't think anyone argues for a distinct treatment of palatovelar+w vs. plain velar+w, since the two series merged in Greek. I don't think we have evidence for the plain velar in this environment anyway except possibly πελεκκον. Relevant question: latin.stackexchange.com/questions/16915/…
    – TKR
    May 8, 2023 at 22:04
  • 1
    That's exactly what people—at the very least Buck and a few Leidenaars—are explicitly arguing, at least in part on the basis of Myc. i-qo being written with qo vs. (secondary) kw being written ko-wo or similar elsewhere (though the example Buck actually gives, te-tu-ko-wo-a₂ (tetukʰwoha, from τεύχω), has aspiration interfering). It's all nonsense.
    – Cairnarvon
    May 8, 2023 at 23:07
  • 2
    @Cairnarvon As far as I know the standard account is that monomorphemic *-ḱw- < -ππ- and presumably the same would hold for *-kw- if we had any examples; it doesn't seem to necessitate any difference in treatment between palatovelars and plain velars, unless there's some example of the latter I'm unaware of that gives something other than -ππ-.
    – TKR
    May 9, 2023 at 0:13
  • 1
    @JD Depends on the OS, but I simply use the Greek Polytonic keyboard on Mac OS.
    – TKR
    May 10, 2023 at 18:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.