On Tue, 30 Oct 2001 16:19:55 +0100, Miguel Carrasquer Vidal
<
mcv@...> wrote:
>So why the PIE perfect with zero grade in the plural?
>I don't know. Maybe the reduplication had something to do with it,
>maybe it's just analogical after the active (but why not go all the
>way and normalize the act. sg. as well?).
Upon reflection: the reduplication, for sure. A perfect pl. like OI.
<ca-kr-úr> *has* e-grade in the pf. plural, just like Hitt. <sekkir>,
the only difference is that the e-grade is in the reduplication,
instead of in the root [*].
This means that the reduplication must originally have been a feature
of the plural forms only (an optional feature, given its absence from
Anatolian, and from a verb like *woid-). This makes good sense, given
that reduplication is universally used as a device to denote things
like iterative Aktionsart, or plurality of either the subject or the
object. It makes no sense as a perfect marker.
[*] So, we must've had (from *gWem- (pf. stem **gwa:m-) "to come",
Hitt. sakhi "I know", sekkir "they knew"):
3sg. **gwá:m-e > *gWóme Skt. (ja-)gá:ma
3pl.A **gwa:m-ént > *gWemé(:)r [Hitt. sekkir]
3pl.B **gwa:gwam-ént > *gWegWmé(:)r Skt. jagmúr
This might also explain why in Vedic the reduplicative vowel is
sometimes /a:/.