On Wed, 04 Apr 2001 05:10:18 , "Glen Gordon"
<
glengordon01@...> wrote:
>>As I said, my current theory is that it comes from the diminutive
>>*-ik-(/*-ih2).
>
>There's nothing internal or external to IE to warrant a "diminutive" *-ik-.
Excuse me? The main PIE diminutive is *-(i)ko- (besides the also
widespread *-(V)lo-).
Greenberg, in "IE and its Closest Relatives", adduces some external
comparanda: Finno-Ugric -k, -ka ~ -kä (dim.); Northern Samoyed -ku
(dim.); Old Turkish -(a)k (dim.), Karaim -ka (fem.), Gagauz -ika
(fem.) [Greenberg's example: qoms^u "neighbour", qoms^u-ika "female
neighbour", but I wonder if this may not be Slavic influence]; Mong.
-[i]ka(n) ~ -[i]ke(n) (dim. and fem.) [e.g. noya(n) "prince" ~
noy-ika(n) "princess"]; Korean -k (dim.); Jap. -ko (suff. of female
names); Gilyak -k (fem.); Eskimo -ax-aq (dim.).
One may add Basque -ko, -s-ko (dim.)
>>So if *-x became lengthening (*h1, in fact), where does *-h2
>>(collective, feminine) come from?
>
>I will reiterate: Consonantal *-x disappears producing vowel lengthening.
>_CONSONANTAL_ *-x disappears. Syllabic *-x continues on as in *-k^ontx
>"-ty". If *x were not syllabic here, we would end up with an ugly *-CCC
>cluster!
So what about feminine *-ih2, *-eh2? Not syllabic, and doesn't just
disappear producing vowel lengthening (*-ih2 gives -ia, not -i:, in
Greek).
=======================
Miguel Carrasquer Vidal
mcv@...