<Language> Re: Caucasus and Kaska

[H. Mark Hubey]

Claire Bowern wrote:

> >If linguistics really took "initial conditions" into realistic
> >consideration, we would have been spared the pouring of much ink!

> >Pat

> Fair enough, but since truly initial conditions are irrecoverable,
> "earliest known stage" will have to do! This is, incidentally, another good
> example of why languages are different from thermodynamic models.

> Claire

Actually, this is a very relevant and interesting post. It is like a
breath of fresh air in a room full of cigar smoke. Whether the earliest
stages are really irrecoverable is not known. And the second statement
is trivially true, but is more relevant to this discussion because it
hides something everybody talks about but rarely comes out in the open.
That is one of the problems with social sciences. The determinism is
masked by the randomness. I say this because standard analytical tools
always use randomness for all forms of uncertainty; errors, real
randomness, chaos, many causes all rolled up into one, etc.
Interestingly enough, thermodynamics is even more useful here because it
really has two faces. IT is also called "statistical physics"! The
classical thermodynamics is the part where the randomness is averaged
out to produce simple deterministic equations. Aha! That is what I have
been trying to point out. Despite all the uncertainty, randomness,
chaos, there is still something we can do. We can create simple and
deterministic models and use them as stepping stones to create more
realistics and more accurate models which definitely (at the cost of
more mathematics) predict the real world much better than the simple
models. Think of Darwinism. What has it ever predicted? Nothing.

Best Regards,
Mark
-==-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
hubeyh at montclair.edu =-=-=-= http://www.csam.montclair.edu/~hubey
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=