Re: Finnish KASKI

On Fri, 28 Mar 2008 16:18:42 +0100, "fournet.arnaud"
<fournet.arnaud@...> wrote:

>
>----- Original Message -----
>From: Miguel Carrasquer Vidal
>>
>>> =========
>
>The reconstruction given by Sammallahti is *kota.
>
>=============
>
>How does *kota account for Hungarian ha:z
>with a long a ?

Abondolo derives that from PFU *kaata (= Sammallahti *kota).

>=============
>>
>>Why did Permic lose the final consonant
>>when it did not in some other words
>>of the same language ?
>
>Because Permic lost the intervocalic stop (*kota > *kua).
>
>========
>Permic :
>n'oma > n'ima
>n'ole "arrow" > n'il
>
>Out of *kota
>we should expect *kia
>
>(assuming this law were right)

If I tally all the words of the pattern //o..a// in
Sammallahti's wordlist (in Sinor 1988), I get:

Komi (Zyryen)
/u/ 19
/o/ 9
/i/ 2
/ë/ 1
/o/ ~ /ë/ 1
/y/ ~ /u/ 1
/y/ ~ /ë/ 1

Udmurt (Votjak)
/u/ 24
/ë/ 1

The regular outcome of /o/ in a-stems is undoubtedly /u/.

The two cases with /i/ in Komi both have a palatal initial
(n'imal, c'ir), which suggests conditioned fronting of the
vowel.


=======================
Miguel Carrasquer Vidal
miguelc@...