Re: Evidence for Nostratic? (was Re: Proto-Uralic?)

En réponse à Rob Haden :


> >How about *septm which is also accented on its syllabic consonant?
>
>OK, something has to give here.  PIE, as currently reconstructed, is a
>mess.  It's like a chimera -- some kind of outrageous form that could never
>have actually existed.  There's just no way to make sense of it all as it
>is right now, therefore the reconstructions must be (partially) incorrect.
The problem is that we just *cannot* consider PIE to be a single instance 
of a language at one time. Reconstruction allows us to get forms of 
different geographical origin (PIE wasn't ever monolithic, no language can 
be, especially when it's illiterate) and from different *times*, without 
always giving us clues of where and when they originate. So reconstructed 
PIE is actually a 3-dimensional (2 dimensions of space and 1 of time) 
mosaic of words and features, which don't always fit together because they 
sometimes never existed together. You must never forget that when you talk 
about PIE, or you will fall in the trap of thinking that PIE is a single 
entity comparable to "Standard French" for instance. It's not. It's a 
necessary incomplete (because you can never completely reconstruct a 
language. Take all the Romance languages and try to reconstruct Latin with 
them, and you will never manage to reconstruct the case system of 
Republican Latin, since it has disappeared without a trace) set of features 
with a margin in both space *and* time. Think of taking all the British 
dialects of English of the last 500 years, pick up different features of 
them all, and look at the result, and you will get an idea of what PIE 
really is. The incoherence is not there because the reconstruction is 
incorrect or meaningless. It's just the best we can do with the tools we 
have, and we must never forget their limitations.

>I'm particularly frustrated with "laryngeals" popping up in every other PIE
>reconstructed word (or so it seems).  I think PIE-ists are just using those
>when they Really Don't Know what the correct form is.  But maybe I'm wrong
>(once again, lol).
You are. I've seen transcribed Hittite texts (they were written in 
Cuneiform, so we can read them easily). The laryngeals in Hittite (it has 
two, transcribed usually "h" and "hh") appear pretty much everywhere, and 
the most important point is that the laryngeal theory of phonology of PIE 
had already appeared (not under this name) before Hittite was recognised as 
an IE language, and the correspondence between the Hittite laryngeals and 
the "sonantic coefficients" that would come to be called laryngeals is 
almost perfect! You just cannot dismiss the laryngeal theory as something 
used "when they Really Don't Know what the correct form is". The 
correspondence between the "sonantic coefficients" (Saussure called them 
that way) reconstructed for purely internal reasons (to explain strange 
alternances) and the Hittite laryngeals is too obvious to be just a 
coincidence (especially since Hittite is an IE language). The laryngeals 
*did* exist, and pretty much everywhere where they are reconstructed (I 
know of only 1 case of a laryngeal put there for purely theoretical 
reasons. It's the case of *H1ed-: the "eat" root. H1 is there only because 
PIE roots are usually biconsonantic, and **ed- - which gives the same 
outcomes in all known IE languages as H1ed- - would be the only known case 
of a monoconsonantic root. So for theoretical reasons, by comparison to 
Semitic languages where words can never begin with a vowel, H1 has been 
added to the root. It's the only case I know of a laryngeal added without 
some strong evidence that it was indeed there). We just have to accept that 
PIE was a pretty guttural language with lots of consonants. So what? it's 
not as if such languages don't exist *now*! look at Caucasian languages for 
instance!

Christophe Grandsire.

http://rainbow.conlang.free.fr

You need a straight mind to invent a twisted conlang.