Re: [tied] Re: Eggs from birds and swift horses (was: the palatal s

On Sat, 14 Feb 2004 14:58:28 +0000, elmeras2000 <jer@...> wrote:

>--- In cybalist@yahoogroups.com, "tgpedersen" <tgpedersen@...>
>wrote:
>
>> I was wondering once if the -s- of *h2wegs- might be the result of
>a
>> back-formation from an s-perfect (not that I know much about it)?
>
>If you mean s-aorist, we may be in business. It seems that the
>sigmatic aorist belongs so often with an sk-present that the
>conclusion is forced upon us that their original functional nuance
>was the same. Now, with the sk-verbs that is known to be inchoative.
>Extrapolating from there, one will think of *H2wegs- as 'begin to
>grow', while *H2ewg- is just 'grow'. That looks nice, but I am not
>sure it is correct. My reservations are as follows:
>
>The sigmatic aorist does not cause the root-vowel to change its
>place in the consonant skeleton. Still, Benveniste has a rule that
>it should: "un thème à l'état I n'admet pas d'élargissement; seul
>l'état II en comporte" (Origines 153). He quotes the s-aorist *prek^-
>s- which to him is from a root *per- forming zero-grade *pr- +
>suffixe -ek^- + élargissement -s-. Now, that is fine, but only
>because the root was really *prek^-, not *perk^- (except if changed
>to that). There is no reference to a root structure *TeRT- which
>forms s-aorist *TReT-s-. Actually the aorist stem of *deyk^- 'point'
>is *de(:)yk^s-, not §*dyé(:)k^s-, cf. Lat. di:xi: and Avest. da:is^,
>imperative do:is^.i:. In fact not a single example in the list of s-
>aorists found in Macdonell's Vedic Grammar (§ 522) or in LIV's index
>shows such a transformation. I have often wondered how Benveniste
>got away with writing all the nonsense he is quotable for.
>
>An alternative identification of the aorist -s- is with the
>reflexive pronoun *s(w)e 'oneself' which core is /s/. The there
>could even be some sense in the fact that *H2ewg- means 'augment,
>make bigger', while *H2wegs- is 'grow bigger', i.e. 'make oneself
>bigger'. The same would apply to *H2elk- and *H2leks- 'ward off,
>defend', which could also quite well be base-verb and reflexive.
>However, these speculations cut back into a pre-PIE morphology we
>cannot control.

The lengthening of the root vowel in the s-aorist suggests to me that the
original shape was *deyk^s > *de:yk^s, with length due to normal
"Szemerényi lengthening", as in the nom.sg.. If the origin was really
*swe, then it had already been reduced to *-s(w) by the time the
lengthening set in.

At the risk of repeating myself, I prefer to explain the *-s as a third
person ending, derived either from [the nominative **su of] *s(w)e (which
was a 3rd. person pronoun before it became a reflexive) or the
demonstrative *so. The original aorist paradigm would have been:

*déik^-m *dik^-més
*dé(:)ik^-s *dik^-té
*dé(:)ik^-s *dik^-é(:)r-s

This was regularized to the root aorist:

*déik^-m *dik^-més/-mé(n)
*déik^-s *dik^-té(r)
*déik^-t *dik^-é(:)r-s

But, based on the 2/3 person sg., it also gave rise to the s-aorist:

*dé:ik^s-m *dé:ik^s-me(s)
*dé:ik^s-s *dé:ik^s-te
*dé:ik^s-t *dé:ik^s-(e:)r

The 3rd. person preterite ending *-s occurs in Hittite and Tocharian (which
don't have an s-aorist), as well as more generally in 3pl. *-e:r < *-er-s,
with Osco-Umbrian variant *-ent-s. For a similar development (3sg. ending
-> preterite ending), cf. the Old Irish t-preterite.

=======================
Miguel Carrasquer Vidal
mcv@...