I found it. Thanks to Richard. I agree with this
statement totally:
------------------------
Something here is not right. I can't imagine Proto-Nostratic speakers
needing so many words for "turning". Obviously the majority can be described
as:
velar stop-vowel-(d/l/r/n)- in other words they could could derived by common
and well-known changes from something like
*kVl-
or
*kWVl-
-----------------------------------
I have been working on these myself. There is something seriously wrong with
the way
things are done. That calls for some kind of an overhaul of the methodology.
In fact, I have been reading and thinking about these for years and I have
come to
certain conclusions myself. Here is a simple way to look at things in the
sense of
"why is this insufficient". Suppose you collect all the roots of various
languages,
as done, for example, by Bomhard (which by the way is excellent). There
is still
a problem. The problem is that although patterns were used to create these
roots,
there are still patterns in the result. There are patterns in patterns and
there is
no reason to call a halt to the search for patterns because all science is
based
on patterns, and none of the sciences have yet come to a halt because none
of
them are really finished.
One day it hit me. Humans used stone tools for everything for hundreds of
thousands of years. Breaking, smashing, digging, clipping, cleaving, slicing,
dicing, drilling holes, etc was done with stone tools. So after I had spent
enough
time looking through dictionaries, including Hittite and Akkadian, I realize
that
this regular sound change business is insufficient. Then I discovered to
my
delight that in Semitic studies they have already given up on it (Saenz-Badillo,
A History of the Hebrew Language). A new model of sound change is needed.
The heuristic is still sound, but it is still a heuristic and it is too simple
to totally
explain the data.
My data, the way I constructed things, tells me that the roots that can be
found easily (because they are the latest) are KR and KL. I see that these
are from KTh and thus using some simple sound changes *th>{w,l,s/sh} etc
some of which I gave already in various groups, I can derive most of these
words, maybe all accross Semitic, IE, Turkic and even NS. I do not claim
that the sound changes took place exactly like that. These are what might
be
called changes based on something like Parsimony Principle, Occam's Razor,
MDL (minimum description length), MEM (maximum entropy method), etc.
In other words they are reasonably concise and precise descriptions of
observed data. Then I saw that I can take this backwards e.g. T>K,
and then further back P>T. The trick is to fit the data to a theory, a
mathematical theory, something I have been working on for a decade or
more. I am convinced that these things done with stone are derived from
words for stone and can be seen to have the same phonetic form accross
language families.
In fact, the root kVr/kVl for "turning" can be also found in the form
tVr. The earliest form for "cut" etc PL root (pilakku, pelekus, BALAG, balta
polat, bilda.) can also be found as PR (part, portion, parathu, parala, parchala..).
It looks like tVr gave rise to kVr but this is just one aspect of a big problem.
What I cannot determine are things like was there a single liquid at one
time,
which was it if it was so, were there two or more, why, was there a single
language, or was there a mixing of 2 or more, etc. These are the things that
interest me, not arguing about silly things regarding IE as if it was the
holy
language of god. The reason everything is claimed to be IE is that it is
the first one deeply studied as a result of a historical accident. It was
constructed as a vacuum-sweeper which collected all the words that could
be
found. I am looking for other principles to decide things, not fight stupid
fights of which I am already disgusted.
Here is the specific problem; Turkic has
Chuvash pěr one [Krueger61:226];
Turkish bir, one. Other dialects might have bIr.
Chuvash pus' head, beginning [Krueger61:227];
Turkish bash, head.
Also bashla to start, to head
Common Turkic barmaq/parmak finger;
Chuvash pu"rne finger [Krueger61:228];
Turkish parmak, finger.
This word and the word for giving (ber) might be related in the sense that
ber might have meant ‘to hold’.
The words for "one", "head", "finger", and words like "to start" all are
basically from the
same root. This root means "first" more or less in IE. I cannot believe that
an IE word
got into Turkic and replaced three words from the Swadesh list. The word
even shows up
in Circassian parma (first).
But it shows up in both the centum and satem branches of IE. I know by now
that
no self-respecting IEanist will ever give up such words. I consider it irrational
but
there are too many of them. It seems to me that they are working against
the rules
which they claim to be upholding. So apparently we are in need of another
general
principle/heuristic because as I said RSC is insufficient.
This problem is a general and abstract one. Problems like these have to be
solved in a general way so that they are applicable accross language families.
So that is a Nostratic problem.
I would welcome a serious discussion, but no insults. After 10+ years of
enduring irrational chauvinism I am short on patience.
H.M. Hubey wrote:
-------- Original Message --------
--- In Nostratica@yahoogroups.com,
"H.M. Hubey" <hubeyh@...> wrote:
> I thought this was a Nostratic
list. It is obvious that there is
at
> least one *kir root. And there
> must be at least one *kwel root
because IEanists say so.
The problem here is that there seem
to be a lot of *different* kVR
roots with a meaning like 'curve'.
See e.g.
http://groups.yahoo.com/group/nostratica/90
.
Richard.
--
M. Hubey
-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o
The only difference between humans and machines is that humans
can be created by unskilled labor. Arthur C. Clarke
/\/\/\/\//\/\/\/\/\/\/ http://www.csam.montclair.edu/~hubey