Re: [tied] Polish G. -ga

On Mon, 13 Jan 2003 02:05:26 +0100, Miguel Carrasquer <mcv@...>
wrote:

>No, that I see as inherited from the feminine marker *ih2/*yah2, more
>or less as follows:
>
>N **tá-ix > *té-h2 > *ta: > ta
>A **tá-ix-m > *té-h2-m > *ta:m > toN
>G **ta-íx-âs > *to-yáh2-os > *toyá:s > toye(N)
> [**ta-ix-ás > *to-yh2-és > *toyyás]
>DL **ta-íx-a(i) > *to-yáh2-i > *toyá:i > toyi
> [**ta-ix-á(i) > *to-yh2-ái > *toyyái]
>IA[**ta-íx-ât > *to-yáh2-ot > *toyá:t]
> **ta-ix-át > *to-yh2-éh1 > *toyyá: > toyo(N)
>
>N **í-ix > *yé-h2 > *ya: > ya
>A **í-ix-m > *yé-h2-m > *ya:m > yoN
>G **a-íx-âs > *e-yáh2-os > *eyá:s > yeye(N)
> [**a-ix-ás > *e-yh2-és > *eyyás]
>DL **a-íx-a(i) > *e-yáh2-i > *eyá:i > yeyi
> [**a-ix-á(i) > *e-yh2-ái > *eyyái]
>IA[**a-íx-ât > *e-yáh2-ot > *eyá:t]
> **a-ix-át > *e-yh2-éh1 > *eyyá: > yeyo(N)

It should be added that only Slavic (and perhaps Latin) preserve(s)
this declensional model in the feminine pronoun. Sanskrit and
Germanic use a morpheme *-s(y)- [contracted from *-sm-ih2- or based on
G. *to-sy-o ~ *tosy-a:s ?] (G. tá-sy-a:s, D. tá-sy-a:i, L. tá-sy-a:m;
G. þi-z-o:s, D. þi-z-ai) (*), and e.g. Lith. shows a normal a:-stem
pattern (G. tõs, D. tái, L. tojè, I. tà). The Slavic pronominal
paradigm is, however, identical to the Sanskrit a:-stem nominal
declension:

G. -a:ya:s < *-oya:s < *-oyáh2os
D. -a:ya:i < *-oya:i < *-oyáh2i
L. -a:ya:m < *-oya:m < *-oyáh2(+m)
I. -aya: < *-oyya: < *-oyh2áh1

(*) the Slavic feminine forms cannot go back to *tosya:s etc., given
the only surviving Slavic masculine genitive form c^eso (c^Iso) <
*kWesyo, which shows that *sy- would have given Slavic -s- here.


=======================
Miguel Carrasquer Vidal
mcv@...