Proto-Quendian (and PIE) phonology (was Re: Q (Caucasian Elf))

Andreas Johansson <and_yo@...> writes:

> Jörg wrote:
[PQ consonant system snipped]

> Some would tell you that Tolkien's PQ also had a few palatalized sounds: t',
> d', k', n' (I'm probably forgetting one or a few) that could contrast
> against mere clusters ty, dy, ky, ny. One of the points of this would be
> that it's easier to understand how Quenya _Solonyeldi_ can correspond to
> Amanya Telerin _Soloneldi_ while Quenya _Vanyar_ corresponds to Amanya
> Telerin _Vaniai_.
These two examples seem to indicate that there were two different
phonemes in play, at least when accepting the Neogrammarian approach;
one of the forms might be irregular, or it is conditioned by the
following vowel:

*nj > ni / _{a, o, u}
*nj > n  / _{e, i}

or anything like that.  Or PQ *nj always goes to *ni, but *ie > e and
*ii > i in the next step.  No need for a contrast between /nj/ and /n'/.

And finally take notice that my PQ is not the same as Tolkien's PQ, and
that the phonology of my PQ is not yet fixed.  In fact, the existence or
non-existence of a palatal series is one of the major open questions in
"my-PQ" phonology.  I think I'll get along without one.  After all, this
is not the protolanguage of Quenya, Amanya Telerin and Sindarin, but of
Nur-ellen, Caledonian, Iverinic and Macaronesian.

But as we are at it, I am sceptical of the palatal series in PIE as
well.  I'd rather say that they were velars, which somehow were missed
by the satemization.  (I am not alone here; Szemerenyi does the same
thing, if I understand him corretly.  He also doesn't have a palatal
series in his reconstruction of PIE.  Of course, he has other
controversial things as well, such as the voiceless aspirated grade
which most others threw out a few decades ago, and only one laryngeal,
namely /h/.)

Note that in Baltic (a geographically marginal satem branch), some words
have /k/ where other satem languages have /s/.  Do they justify
reconstructing yet another series of stops between palatal and velar?  I
don't think so.  Sound changes are not all that perfectly regular in
reality.  For an example, look at the way the High German consonant
shift peters out in the Rhinelands: just about each single word affected
has its own shifted/unshifted isogloss (though most of these isoglosses
tend to cluster around a handful of lines).

There's another problem with the traditional reconstruction: it would
indicate that palatals had been much more frequent than velars in PIE,
which seems odd.  Assuming an irregularity in the satemization clears
this problem.

Jörg.