--- In Nostratica@yahoogroups.com, "H.M. Hubey" <hubeyh@...> wrote:

>
>
> Richard Wordingham wrote:
>
> >
> > > >
> > > It would if p>t>k occcured.
> >
> > Unconditioned t > k is very rare. In the examples I can think of

(all

> > Polynesian, some still in progress), k was missing (also

unusual), e.g.

> > because of a shift k > h. It is associated with very small

consonant

> > inventories.
>
> If the RSC is true, all the t-words would disappear leaving only k-

words

> so it would not be
> possible to know anyway. There is also the possibility of p>k>t. I
> discuss this in my article
> in the J. of Quant. Ling. Assoc. It is also in my book. And I do

find

> cases of these happening.

One of the advantages of the Austronesian languages for studying
sound changes is that it is such a large group. A change t > k shows
up in comparison with other languages, and if you had, say c~t~k
collapsing to t~k~k, it should show up by comparison with a large
number of unaffected languages. (I know, I ought to write t~k, but I
let you call your bags - or were they sequences? - sets.)

What are your accepted p>k>t and p>t>k examples?

> > You're thinking in terms of continual borrowing as the
> > explanation of the Nostratic group, rather than common descent.
> > Do you also
> > see that as the explanation of the Indo-European group?

> No, I am thinking more like mixing of two language families, and an
> approximation of what
> happened as two phases: the initial mixing phase (chaotic,

turbulent),

> and the secondary phase,
> (more laminar) in which a lot of the RSC took place, and enough

time

> passed to produce
> more regular things; a kind of a relaxation phase.

Sounds more like the Melanesian Austronesian languages. Any
Austronesianists on the list?

<snip>

> > > Turkic which stretches from the Pacific to the Adriatic.

> > Reaching the Adriatic is fairly recent.

> Large numbers of them must have existed for a long period of time
> in order to have spread
> out and not disappeared. Supernova-ing is a rare event. In any

case

> they were in the Asian region for a long time.

The impact of 'supernovas' is quite wide, and there have been quite a
few in historical times:

Big ones, in historical order:
Latin, Arabic, Spanish, English

Moderate ones:
Aramaic, Turkish, Russian

I don't know how to count the Hellenistic expansion of Greek. I
don't think the Greek and Phoenician coastal colonisations count as
supernovas.

On the edge of history, i.e. I think poorly documented:

Chinese (the biggest of them all), Slavic, S.W. Tai (denied by the
Thais)

And what happened to the Iranians of the steppes? Were they
swallowed up in another Turkic expansion? (There are claims that
some clans were absorbed by the Mongols).

Unrecorded, but clear:
Austronesian, Bantu

Eskimo-Aleut looks impressive on Mercator's projection!

It's not surprising if people use this model for Afro-Asian, Indo-
European (2 waves - agricultural and steppe, and should we count an
Indian expansion as well?), Austro-Asiatic and a prehistoric Tai
expansion.

> > What sort of Mongoloids? Northern or Southern? If they're

Southern

> > Mongoloid, they'll obvously think in terms of Austro-Asiatic

first. Any

> > chance of Central Asian contacts?
>
>
> I am not sure. I have to read the book again. If they were
> Austra-Asiatic they would have been
> found in the south of India, not north.

Witzel has them coming up the Ganges as well. But it's the North West
of India that matters in this case.

Richard.