Re: Greek psephas/knephas/dnophos/zophos: linked?

--- In cybalist@yahoogroups.com, "Torsten" <tgpedersen@...> wrote:

>
> > According to R.S.P. BEEKES "PRE-GREEK. A LANGUAGE RECONSTRUCTED",
> > there seems to be a relationship between many Greek words meaning
> > "darkness":
> > psephas
> > knephas
> > dnophos
> > zophos
>
> also gnophos
>
> > Beekes tried to relate knephas/dnophos < *kdnophos, but he couldnt
> > connect all four names.
> >
> > psephas, cf. Sansk ksap- < *kWseph-
> >

We've got cognates in Indo-Iranian *k^sep- and Hittite ispant- 'night'. But as Greek -ph- corresponds to stop series I instead of series III, pséphas must be a "Pelasgian" (a variety of Thracian) loanword, as in that language series I was voiceless aspirated as in Germanic and Armenian.

There's also the Altaic cognate *dz^ipHu 'evening, darkness' (Tungusic *dz^ip-ku 'to dusk', Japonic *dupu 'evening'), whose sibilant would explain Greek z- in zóphos.

Comparison with NEC *h\nitts\wV 'night, evening' (> IE *nekW-t- 'night') suggests the labial stop is the result of the reduction of the affricate+labial glide cluster. This way, Greek knéphas could be straightforwardly from the NEC protoform.

> *dhnebh- < *dhnembh- < *tŋ-áN- < *taŋ´-aN-? [Pokorny's *tem- "darkness"]
> (*nembh- is then no 'Kontaminationsform')
>

This reminds me of Altaic *tHjú:ni  'night, yesterday, dark'.

> which also takes care of the odd Lithuanian form with d-
>

IMHO This would be explained as a denasalization *n- > *d-.

> Semantically "darkness" and "cloud" match.
>

That's right, but 'night, evening' has got preference.