Re: Armenian, Sumerian, Burushaski, and Turkic languages

In article <1181627608.741509.161940@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>,
Franz Gnaedinger <frgn@xxxxxxxxxxx> wrote:

On Jun 12, 6:45 am, Nathan Sanders <nsand...@xxxxxxxxxxxx> wrote:

Likewise, the comparative method excludes 1=2, just as it should. Why
would we want the comparative method to allow for results that are
mathematically impossible?

The comparative method says that for example fungor
and fungus have no common root.

No it doesn't. It says that we cannot identify a common root. Those
are two completely different claims, and if you can't understand the
difference, you have no business discussing historical linguistics.

There _are_ unoutspoken assumtions, for
example about the length of early words. And for these
assumptions you got no proof, they are conveniences.

There are no assumptions about actual word lengths in the comparative.
However, there is a lack of assumption about particular word lengths.
Again, a crucial difference that you seem not to have the capacity to
grasp.

Nothing reliable about specific words can be said about that point in
time. If you imagine that people used the word [ak] to refer to water
40,000 years ago, that is a complete fabrication on your part, with
absolutely no scientifically sound supporting evidence.

What I ask for is a theory of early and very early
language. It's missing.

There are plenty of robust, scientific, well-supported theories about
early language. None of them can say anything about specific words,
however. It is a mathematical impossibility given our current
evidence.

We certainly cannot assume that "early" language consisted only of
very short words and "figurative compounds".

Using the word "certainly" speaks of an assumption.
Where is your evidence?

My evidence comes from logic. If we have two things (human physiology
from now and 10,000 years ago) that are identical in every relevant,
measurable way, then without further evidence, we cannot assume that
some particular property (the underlying structure of human language)
is different.

The null hypothesis is that humans with the same size brains have
language within the same range of typology. Without evidence to the
contrary, the null hypothesis must be maintained, and you have *no*
evidence whatsoever to reject the null hypothesis.

"Null hypothesis" sounds impressive, but is nothing
else than a basic assumption. It contradicts with
another feature of language, namely that language
mirrors life, and in humans the level of technology.

The underlying structure of human language does not "mirror life".
The complexity or simplicity of a culture has nothing at all to do
with the complexity or simplicity of the underlying structure of the
language it uses. Portion of English grammar have become more simple
over time: we've lost noun cases and most person-number distinctions
in verb inflections. Meanwhile, some far more primitive cultures have
incredibly complex grammars, with noun classifiers, highly inflected
verbal systems with numerous tense and aspect distinctions,
multi-tiered honorifics, etc.

I formulated this relation back in 1974/75, and it
seems no one else has done the same, whether
before me, nor since then.

Because it's an idiotic relation. It doesn't exist in the real world.
It is contradicted by actual evidence from real human languages and
cultures. Anyone who has taken even the most basic introductory
linguistics course would know that your "relation" is utter bull***,
suitable for fiction writing, but not for basis of scientific inquiry.

There just isn't any reason except uninformed fantasy to claim that
the language of humans from any time period with brains the same size
as ours had consistently shorter words or was some bizarre form of
metaphoric communication like the language of Star Trek's Tamarians.

Neanderthals had bigger brains than we have,
so they must have used longer words than we do ...

Another idiotic statement, and nothing that can be derived from what I
have said.

Nathan

--
Nathan Sanders
Linguistics Program
Williams College
http://wso.williams.edu/~nsanders/
.

Quantcast