On Sun, 20 Apr 2003 14:34:34 +0000, Glen Gordon
<
glengordon01@...> wrote:
>
>Miguel:
>>The thematic vowel _itself_ does of course behave most peculiarly
>>with regards to the zero grade rule.
>
>Of course it does. The development of the thematic vowel started
>only in the Late IE stage. It stems from *& as I've said,
>lengthening before voiced segments. This qualitative ablaut
>therefore is different in origin and more recent than that
>underlying perfect and stative verb forms. It also POSTDATES
>the loss of unstressed vowels that caused quantitative ablaut
>in Mid IE, so it does not operate under quantitative ablaut.
The point is that the thematic vowel has to survive quantitive ablaut
in order for any further developments to be able to occur later. So
the mere existence of the thematic vowel must and does affect the
formulation of the zero-grade rule.
Let's a 3sg. thematic verbal form like *bhéret. If the zero grade
rule is formulated as "unstressed /e/ (or whatever its predecessor,
say /a/) is dropped", we would expect **bhár-a-t > *bhér-t, or
*bhar-á-t > *bhré-t. So that won't do.
If we reformulate it as: "unstressed non-morpheme final /e/ (/a/) is
dropped", we get *bhár-a-t > *bhér-e-t (OK) or *bhar-á-t > *bhr-é-t
(not OK, but OK for the tudáti type).
Yet another possibility is the vrddhi solution (1. "addition of the
thematic vowel causes lengthening of the root vowel" [proposed vrddhi
rule], 2. "zero grade affects unstressed, non-thematic /e/, and
shortens long pretonic vowels to /e/ [full formulation of the
zero-grade rule]; 3. accent is retracted to the first full vowel
[Jens' initial accent rule]). This produces: *bhá:r-a-t > *bhór-e-t
(not OK, but OK for types such as adjetival *bhór-o-s) and *bha:r-á-t
> *bhér-e-t (OK).
=======================
Miguel Carrasquer Vidal
mcv@...