On Thu, 29 Nov 2001 03:39:17, "Glen Gordon" <glengordon01@...>
wrote:

>
>Alright, Miguel, *toi. I'll give you that one. However, even so,
>**tos simply looks too much like an animate nominative and *-yo
>is still necessary to make an OBVIOUS-looking genitive.

That's circular. Were it not for the fact that it's the familiar
pronominal (and from there, o-stem) gen. ending, there's would be
nothing obviously genitive about *-osyo. In fact, it looks so
"obviously" genitive that Hittite dropped it in the o-stems and
replaced it by *-el in the pronouns, that Slavic dropped it in the
pronouns (except c^eso < *kwe-s(y)o) and replaced it by Abl. *-o:t in
the o-stems, and that Latin had to add -s to it in e.g. <eius> (*esyo
+ s).

>Since
>most people accept that *-yo HAS been secondarily affixed to
>the genitive, your arguement is pointless anyways because it
>fights against common consensus in favour of something ad hoc.

There is no consensus about *-osyo. Affixing of rel. pron. *-yo is
one of the theories on the table, but, as Szemerényi puts it, this
"remains for the time being disputable" (although he personally
favours it). Since genitives usually develop out of adjectives, it is
interesting to note that Luwian has no genitive, and uses the i-stem
adjective -assi-s to make pseudo-genitives. Note also that the Etr.
gen. in -s can be reconstructed as *-si, on the basis of the compound
endings Gen+Gen=Abl and Gen+Loc=Dat:

a-stems e-stems u-stems i-stems C-stems
s-Gen -a-s -e-s -u-s -i-s -C-s < [*-si > -s]
s-Abl -e-s -e-is -u-is -i-s -C-s < [*-si-si > -(s)is]
s-Dat -a-si -- -u-si -- -C-si < [*-si-i > -si]
l-Gen -a-l -e-l -u-l -- -C-l < [*-la > -l]
l-Abl -a-las -- -- -- -- < [*-la-si > *-las]
l-Dat -a-le -- -- -- -- < [*-la-i > *-le]

This justifies an analysis of the form *-osyo as *-osy-o (Luw.
-assi-a) instead of *-os-yo.

>>root in zero grade: *wlkW-os, *tud-é-ti
>
>I've already explained *wlkWos as being a genitival construct
>and as having original vocalism but unoriginal accent caused
>by a later regularisation of accent onto the initial syllable.

Ok.

> present
> -------
> singular plural
> 1p *beréme *beremes
> 2p *berése *berétes
> 3p *beréhe *beréne
>
> aorist
> -------
> singular plural
> 1p *beréme *beremése
> 2p *berése *beretése
> 3p *beréhe *berenéhe
>
>So once the vowels dropped due to heavy stress, this left
>an accent always on the final. Hence, Early Late IE:
>
> singular plural
> 1p *berém *beremés
> 2p *berés *beretés
> 3p *beré:t *beré:nt
>
>What's interesting is that this theory predicts original
>lengthening in the 3ps due to loss of *h (*H1). The lengthening
>could be easily leveled out, of course. Totally neatoh!

I'm puzzled. What has any of this to do with the real IE aorist?

We have the athematic root aorist, with the root in full grade/zero
grade (I'll use the root *bher- for illustrative purposes, not meaning
to imply that it all these formations actually were made from it):

*bhér-m, *bhér-s, *bhér-t
*bhr-mé, *bhr-té, *bhr-ér

We have the thematic root aorist, with root in zero grade:

*bhr-ó-m, *bhr-é-s, *bhr-é-t
*bhr-ó-me, *bhr-é-te, *bhr-ó-nt

We have the athematic s-aorist, with root in lengthened grade
(originally only in 2+3 sg., by Szemerényi's rule):

*bhér-s-m, *bhé:r-s-s, *bhé:r-s-t
*bhér-s-me, *bhér-s-te, *bhér-s-er

And we have the thematic s-aorist, with root in zero grade:

*bhr-s-ó-m, *bhr-s-é-s, *bhr-s-é-t
*bhr-s-ó-me, *bhr-s-é-te, *bhr-s-ónt

>>root in e-grade: *bhér-e-ti
>
>Again, *bhereti has original vocalism but unoriginal accent
>due to thematic regularization. The first *-e-, originally
>unaccented in the 3ps plural, avoided loss because, as in many
>other cases (ex. *po:t/*pedos), it has been paradigmatically
>strengthened.

You're ignoring the whole problem. There is no reason for
paradigmatic strengthening in a CeR root like *bher-. In fact, zero
grade *bhr- is common enough. If the root was unaccented, it should
have zero grade (like it has in the tudati-presents and the thematic
aorists). If the theme vowel was unaccented, it should have
disappeared. Neither has happened in the *bhereti-presents or in the
conjunctive. That is the central problem, and you do not address it.

The solution to the problem is simply that the thematic vowel was, in
these cases, a separate word (a pronoun) with its own accent, which
was univerbated only after the working of zero grade.

>>root in long grade: *bho:r-os, *swe:kur-os
>
>The development of vrrdhi has nothing to do with accent.
>It could be because of there being an underlying monosyllabic
>root (automatically lengthened in Mid IE) or because of loss
>of *h (*H1), *x (*H2) or *xW (*H3). Accent simply does not
>cause vrddhi.

The thematic vowel does. The question is why?

>>Because the vocative is endingless and shows the bare stem *ek^w-e. The
>>thematic vowel is *-e in absolute final position.
>
>I realize this. However, the vowel could easily be modified
>due to the locative *e, nonetheless. That would explain why
>it doesn't show up in athematic stems... because it isn't really
>an ending. It's just a misanalysis of the original final *-o
>as a locative *-e.

No, it's the regular development of the thematic vowel in final
position. Vocative *-e, Imperative 2sg. *-e.


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