Re: [tied] PIE *h3 and PPIE **n

On Wed, 13 Dec 2000 05:16:41 , "Glen Gordon"
<glengordon01@...> wrote:

>>But an optional (C) [i.e. *-n(C) > *-r(C)] wouldn't hurt >(considering
>>Ved. ásrk "blood" or yákrt "liver").
>
>I will say again, -k (as well as -t) are secondarily added. The
>reconstructed IE forms for "blood" and "liver" both end in *-r and are given
>no ending, Miguel. Get out a Hittite dictionary. You will see no -k's or
>-t's attached. This doesn't exist outside IndoIranian.

That's a curious statement from someone who has just explained *sáxwl,
*sxwéns as analogical after the -r/-n- heteroclitics. I'd say Latin
<sanguis> proves that the <-k> in Vedic is *not* secondary, just like
Greek <he:par>, <he:patos> proves the <-t> in <yákrt> to be original.

>>Toch A Toch B
>>
>>active present
>>-m -u, -w
>>-t -t(o)
>>-s. -m.
>>-mäs -m(o)
>>-c -cer
>>-ñc -m.
>>[...]
>>active preterit
>>-wa: -wa
>>-s.t -sta
>>-0 -0
>>-mäs -m(o)
>>-s -s(o)
>>-r -r, -re
>
>Right, the 3pp active preterite /-r/ could be a potential catalyst for the
>purely TochB form -cer.

Don't be silly. Perfect *-r(o) doesn't even catalyse an -r in the
perfect 2pl. (Toch. *-s), so why should it influence the imperfect?

It was more likely the old imperfect/aorist/optative ending *-e:r
(later replaced by *-nt, but still seen in Hittite -e:r, and as Vedic
-ur in some impf./aor. and all opt. forms) that influenced the 2pl. in
proto-Tocharian *-te:r, before it was supplanted itself by *-nt. The
Toch. imperfect derives, at least in part, from the PIE optative.

>Is *-te:r your own reconstruction or is it actually
>reconstructed by experts? Where else is this *-te:r attested? (Probably
>Armenian again :P)

Toch. B /e/ has two origins: PIE *o and PIE *e:. The difference is
that *o doesn't palatalize, and *e: does. So -cer can only be <
*-te:r.

>We can't even attribute a common -cer form to the
>Tocharian branch

As I explained, we can. Toch. A has lost the Common Toch. imperfect
plural endings *-m(y)ä, *-tyer, *-0.

>so why must we overturn standard IndoEuropean
>reconstructions

There is no standard IE reconstruction that accounts for Toch -cer.
There is no "standard" PIE reconstruction for the active plural at
all, although there's nothing really unorthodox about *-men(i),
*-ten(i). It's attested in Anatolian.

>and modify a working Heteroclitic Rule just because of one
>deviant form.

What are you talking about? I'm trying *not* to modify the
"Heteroclitic Rule" by explaining an exception like Hitt. -ten and
pointing out an isolated form that might reflect lautgesetzlich *-ter.

>Honestly, I don't understand your reasoning.

Ditto.

>>Tocharian A <se>, B. <soy> (< *soius < *suius). Maybe Armenian ><ustr>
>>(influenced by <dustr> "daughter").
>
>Ignoring /ustr/ for many reasons, how can we derive *-u- out of /se/ &
>/soy/? And surely you would want it to derive from **suxyus, if anything,
>not **suius. There is no IE verb *sui-, Miguel! The verb is *seux-. Again,
>there is no justification for the *-y- in such a senseless form as **suxyus.

*soius and *suius are pre-Tocharian forms (PIE *u > *(w)ä in Toch., so
there's no need to mark the length on *u:). They show that in
Tocharian we have dissimilation, as in Greek, but working the other
way (*su:ius > hu:ios; *su:ius > *soius > *seywä). Tocharian A
clearly shows that the word was an u-stem: NA <se>, G. <seyo>, pl. N.
<sewa:ñ>, A. <sewa:s>. See D.Q. Adams "Tocharian" 1988, p. 127.

>>>There is no **yem- like you expect with your "dual-state" rule.
>>
>>There is.
>
>Perhaps in a conlang dictionary or amongst friends at a Star Trek convention
>but we're talking about competantly reconstructed PIE roots. I have never
>seen this root reconstructed anywhere, so it must have arisen from your own
>flightful whim.

See IEW p. 505 *iem- "halten, zusammenhalten etc." for some Sanskrit
forms. The Slavic forms are listed under *em-.

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