On Tue, 30 Oct 2001 10:49:23 -0000, "Sergejus Tarasovas"
<
S.Tarasovas@...> wrote:
>A question to Piotr.
>In your exposé on the PIE verb
>(www.geocities.com/caraculiambro/Caraculiambro/Verbs.html) you write:
>
>'For example, the root *bHer- `carry' is inherently durative, so the
>preterite *bhér-e-t (corresponding to the present *bHér-e-ti `he
>carries') means `he was carrying'. To express a completed action, one
>would use a specially marked aorist form, in this case *bHé:r-s-t `he
>carried, lifted' (a so-called sigmatic aorist). ... gWHen- ... aorist
>*gWHé:n-s-.'
>
>You don't express your opinion on the reasons of the root vowel
>lenghthening in the sigmatic aorist 3sg. Merely phonetic lenghthening
>of a stressed /e/ before a sonorant or glide in closed syllables?
>Additional morphonemic marker of the sigmatic aorist? Anything else?
That's funny, I was just (re-)reading Szemerényi on this question. A
short summary of what he says: even though the lengthening does not
take place in Greek, and some have said that Greek maintains the
original state of affairs, the evidence from Old Indic, Latin, OCS,
etc. warrants reconstructing the (athematic) sigmatic aorist with
lengthened grade, at least originally in the active singular. The
reason for the lengthening is sought in "Szemerényi's law" (i.e. the
"nominative lengthening" caused by *-s), applied to word-inner
position. Because Szemerényi sees this as a compensatory lengthening,
he has to go from *bher-s-m [through *bherr-m ?] to *bhe:r-m and
analogically back to *bhe:r-s-m. Sz. does not explain why the
soundlaw failed to work outside the act. sg., but presumably that's
because the middle and subjunctive forms are syllabic, which 2sg. -s
and 3sg. -t are not (Szemerényi does not mention the fact that in Old
Indic, the 1sg. act. does never have vr.ddhi in the injunctive).
My conclusion would be that the lengthening originates in 2 & 3 sg.
active only: *bher-s-s and *bher-s-t becoming *bhe:rs and *bhe:rst.
From there, it spread to 1sg. act. (*bhersm. > *bhe:rsm.) and the
active plural (*bhe:rsme, *bhe:rste, *bhe:rsr.). In Greek, the
vrddhied 2 & 3 sg. (plus whatever they had analogically attracted)
were suppressed. In OCS, we have the strange situation that the
paradigm shows vrddhied forms everywhere *but* in the 2 & 3 sg.
(ne^sU, nese, nese, ne^somU, ne^ste, ne^se,), as these forms have been
substituted by thematic root-aorists (nese < *nes-e-s, *nes-e-t).