--- In
cybalist@yahoogroups.com, "stlatos" <stlatos@...> wrote:
> retention of e
>
> *gWexYtu+ > getu 'resin'
>
>
> e > a ?
>
> *dekYm, > *dakYa (sim. tasn) > dOkO
>
> *gYen-xY+ > gOn- 'give birth' (or ablaut?)
>
> kOstaar / kOsteer 'lovely, pleasant'
It's possible that only the short e obligatorily became a. A
completely optional (favoring e>a) change can't be ruled out.
Whatever the extent of the change, the fact that both e, E and a, O
occur in the "centum" words means that, whatever else, they were not
borrowed into a language that only had a and thus substituted a for e.
If the language with KY > K did change e > a in at least some
positions, a close relation with the IE group that also changed e > a
seems likely. I therefore wouldn't consider it that of a "lost tribe"
that went on another migration, possibly at a different time, separate
from Indic but a relative in its proper place as much as the other
uncertainly classified languages in the region (Kashmiri, Dardic,
Nuristani).
Therefore, if those languages show similar changes from regular
rules matching the "centum" words, their relation would be more likely.
For e vs. a there's not much evidence, though some of it could have
been obscured by umlaut, but possibly Khowar:
*gYhel+ > zèhc^ 'yellow'
*teplo+ > *petlo+ > pèt.s. 'hot'
with the same treatment of tR as:
*petr.o+ > *patsr.à+ > pòt.s. 'feather', Koh pAl0 'leaf', Skt pattra-
'wing, feather, leaf'
For retention of K:
*kYatr.o+ > kOtrO 'fight', Kho kHot. t.s.ok- 'fight' v.
*kYLep+ > kOlpin.O 'hide' v., *kYLepto> Kho kHòs^t 'hidden', kHos^téy-
'hide' v.
for change of syl.-final p>f>s^ before dental compare:
*kwapno+ > kHus^ùn 'smoke'
It isn't likely that all kY > kH in Khowar. Aspiration can be
caused by following r/w/y; the dissapearance of r after a fric. in
pòt.s. and
*tr.ekWtlo+ > *tsr.attrà+ > *tsHàtr > tsHàtur 'spindle for cotton'
suggest something like:
kYatr.o . kYLepto . petr.o . tr.ekWtlo . kYr,Ngo
kYatr.a . kYLapta . patr.a . tr.akWtla . kYr,Nga
kYatr.a . kYLapta . patr.a . tr.attla .. kYr,Nga
kYatr.a . kYLapta . patr.a . tr.attr.a . kYr,Nga
kYat.r.a kYLapta . pat.r.a tr.attr.a . kYr,Nga
kxYat.r.a kxYLapta pat.r.a tr.attr.a . kxYr,Nga
kxYr.at.a kxYLapta pat.r.a tr.attr.a . kxYr,Nga
kxr.at.a kxLapta . pat.r.a tr.attr.a . kxYr,Nga
kxHr.at.a kxHLapta pat.Hr.a tHr.attr.a kxYr,Nga
kxHrat.a kxHLapta pat.Hra tHrattra .. kxYr,Nga
kxHrat.a kxHLapta pats.Hra tsHrattra . kxYr,Nga
kxHrat.a kxHLapta pats.Hra tsHrattra . tsYr,Nga
kxHrat.a kxHLapta pats.Hra tsHrattra . ts^r,Nga
kxHat.a . kxHapta . pats.Ha tsHattra .. ts^r,Nga
kxHat.a . kxHapta . pats.Ha tsHattra .. ts^ruNga
etc.
e/o > a
k() > t before t
tl > tr.
t > t. before r.-syllabic
kY > kxY
met. kxYat.r.a > kxYr.at.a
kxY > kx before R-syllabic
obstruent > aspirated before R-syllabic
r. > r
t. > t.s. before R-syllabic
kxY > tsY
tsY > ts^
R-syllabic > 0 after fric.
R, > Ru before CC
This is also similar to Albanian which doesn't palatalize before r/l.