[tied] Re: PIE voiceless aspirates

--- In cybalist@yahoogroups.com, "Patrick Ryan" <proto-language@...>
wrote:

> --- In cybalist@yahoogroups.com, "david_russell_watson"

<liberty@...> wrote:

>
> > Well just to start off, how do you explain the failure
> > of your supposed *kH to produce a palatalized counter-
> > part after the second palatalization?
> >
> > If it was as most nowadays believe, then a laryngeal
> > stood between *k and the following front vowel and
> > so naturally blocked the latter's palatalizing effect.
> > If on the other hand it was a you say, and *kH was
> > a unit phoneme, then we should expected it to have
> > produced a palatalized counterpart to k at the time
> > of the second palatalization.
>
> pPIE *kxi became PIE *k^hV; pPIE *kxa became PIE *khV; pPIE
> *kxu became PIE *khV.

Well putting aside for the moment the fact that nobody
besides yourself posits such changes, they are still
irrelevant to my question, because you suppose them to
antedate PIE, which of course would have them antedate
the second palatalization.

> So far as I can see, there could be no _second_ palatalization

The second palatalization is a well known fact, that
absolutely no Indo-Europeanist would try to deny.

> because, in Old Indian, all *V had become <a>, which could not
> palatalize.

But of course the second palatalization preceded that
change, and is why we have both k and c in Sanskrit
as reflexes of Satem *k. Is this truly the first time
you've heard of it?

> Of course, *kh newly brought into contact with *y
> could palatalize.

Where did it do that?

> I think you mean "if *kH was a unit phoneme, then we should
> have expected it to have produced a palatalized counterpart
> to _*kh_..."

Yes, that's right.

> Also, if <kh> was really /k/ + /x/, palatalization, if it had
> occurred, would easily have produced /k/ + /ç/, which I use
> when I pronounce <cute>.

No, and here again you require me to expound on the
nature of affricates, so don't try to put the blame
for the digression on me again later.

When an affricate phoneme is either fronted or backed
under whatever conditions, the stop component and the
fricative component do so together; they are always
homorganic or they cease to be an affricate. If your
supposed *kx were truly a single phoneme, then after
palatalization nothing other than *t­ç (in which <t>
stands for a palatal stop) could have ever been the
result. Only *k + *x (a sequence of _two_ phonemes)
could after palatalization ever produce anything like
*kç.

So you see it is you who are trying to have your cake
and eat it too, by having affricates act as single
phonemes in one place, but then acting as a sequence
of two in another, completely according to the needs
of your theory at that moment.

> > Now do not try to claim Sanskrit ch as the palatalized
> > counterpart of Sanskrit kh, because ch is a geminate
> > (cch), whether always written as such or not, and came
> > from PIE *sk.
>
> Fine. PIE *k^he becames <S> in Old Indian.

May we see an example of such a correspondence, and
from Indo-European please, not Nostratic?

> > For this reason Sanskrit kh cannot have been a unit
> > phoneme before the time of the second palatalization.
>
> No second palatalization.

Well it's simply ridiculous to try to deny the second
palatalization. Do you deny that Satem *kekore resulted
in Sanskrit cakara? If you do not, then what do you
call the stage in which *k before a front vowel fronted
to something eventually resulting in an affricate in
Sanskrit and Iranian?

> > No, I'm merely saying that the digression on syllable
> > weight and the nature of affricates was not intended
> > as proof, as I thought that I already stated.
>
> Then it is so much empty theory that contributes nothing to the
> resolution of the question.

It's not empty theory, and it contributes to your general
education on the matter of syllable division in PIE, if
not directly to the matter of the voiceless aspirates.

> > Quite easily: the fact that (1) VOICELESS ASPIRATES
> > AROSE OUT OF COMBINATIONS IN PIE OF STOP + LARYNGEAL,
> > leads logically to the conclusion, based upon what
> > is known of PIE metrics, that (2) A VOWEL PRECEDING
> > SUCH A CLUSTER WAS ORIGINALLY LONG BY POSITION, as a
> > vowel preceding _any_ two consonants is supposed to
> > have been.
>
> So what? No bearing on the question.

Things come up in the course of discussion, and in fact
it's more often you who sets the like in motion than
myself.

> > You have misunderstood me as offering (2) as proof
> > of (1), while I have never been doing anything but
> > offering (1) as proof of (2), and so there's nothing
> > at all to reconcile.
>
> Excuse me but #1 is no proof of #2; it is not even relevant.

Of course it is; where's your logic? If in PIE a vowel
was long by position before two consonants (which I'd
have thought a well known fact), and the antecedent of
Sanskrit th was in PIE *t + *X, then a vowel falling
before that cluster would have been long.

> > I will now say still again: for evidence of (1), READ
> > THE BOOK.
>
> If the best you can offer is your "proof" from a second
> palatalization, then no good proof exists.

The second palatalization, which explains how both k
and c in Sanskrit are reflexes of Satem *k, the choice
depending on whether a front vowel originally followed
or not (excluding analogical considerations for the
moment), is one of the most basic facts in all of Indo-
European studies, and I simply refuse to try to explain
it to you in addition to everything else.

> If your references had anything worthwhile to contribute,
> you would already have quoted it.

Back to your old tactics, I see. You can spit and kick
all the dirt you want on a piece of gold, but it will
still be gold.

In fact what you're really doing is shaking your fist
at reality for not conforming to your theories, and
shooting me for no more than being the messenger.

David