On Sat, 06 Mar 2004 11:50:18 +0000, tjodrikperson
<
t9t10@...> wrote:
>> 2sg. *h1i-h1ói-tHa > *i:áytha > iyétha
>> 3sg. *h1i-h1óy-e > *i:á:ya > iyá:ya
>> 3pl. *h1i-h1y-r's > i:yúr
>
>Except that the pre-Vedic forms would have a short *i-, not *i:-,
>since the *h1 drops between vowels without lengthening.
I opted above for an interpretation where we *have*
lengthening, and the /i:/ (= /ii/) becomes /iy/ before a
vowel. Another interpretation is that H remained as glottal
stop or hiatus into Proto-Indo-Aryan (3sg. *i'á:ya), and was
replaced in this case by a palatal glide in Vedic.
The behaviour of a laryngeal between vowels (including /i/
and /u/) is a thorny subject, where different solutions
apply in different contexts. Take for example the root
*reh1-i-, where Vedic has deletion if the /i/ in the acc.sg.
ra:m (*reh1i-m > *reh1m, as in Latin re:s, re:m), and
lengthening [not necessarily caused by the laryngeal] in the
nom.pl. *réh1-ey-es > *ré:-yes > rá:yas).
=======================
Miguel Carrasquer Vidal
mcv@...