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

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

>
> No, it is because I'm sticking to logic. The fact that a set of
> nasal-initial words appears in Altaic or Kartvelian on the one hand
> and in a denasalized form in IE does not necessarily mean that they
> appear in some 'paleo-IE' dialect, since they could all have been
> borrowed from a now lost language, being denasalized when borrowed
> by PIE and not when when borrowed by Altaic and Kartvelian.
>
> > This is precisely my point: your "PIE" is precisely the paleo-IE
> > dialect where this word became denasalized. As I said many times,
> > IMHO the so-called "PIE roots" don't belong to a single language but
> > come from several paleo-dialects.
>
> As long as you haven't defined which IE languages this 'paleo-IE dialect' fed words tp, it is unnecessary by Occam.
>

As I said before, this paleo-IE dialect is part of IE-ists' "PIE", so its outputs can be virtually found in all the historical IE languages. You focus on the outputs while I focus on the inputs.
 

> How do you explaim then that the "fog" word is denasalized only in
> Lithuanian, but your examples (presumably) are denasalized in many
>  more IE languages?
>
> > The fact the nasal survived to denasalization makes me think the
> > word had originally a laryngeal at word-initial: *Hn- (clusters
> > *Hn-/*Hr- are rather common in Proto-NEC, as for example in the
> > 'night' word), then lost in "PIE".
>
> But how would *Hn-/*Hr- explain the d- of debesìs?
>

This *H- would have acted as a kind of containing dam, effectively preventing denasalization from occurring everywhere except in Baltic (like a faulty condom, so to speak). However, I must admit this is only a hypothesis, because in the case of the 'fog' word I don't have an etymology obtained by means of external comparanda.

> > Remember that besides *nebh- we've also got *ºnbh- (e.g. Greek
> > aphrós 'foam' < *ºnbh-r-o-).
>
> But we don't need a laryngeal for getting a- from zero-grade *n.-.
>

Oh! Should I have warned you not to confuse a *real* laryngeal consonant with those *segments* called "laryngeals" by IE-ists?