On Tue, 18 May 2004 12:42:46 +0000, elmeras2000
<
jer@...> wrote:
>--- In cybalist@yahoogroups.com, Miguel Carrasquer <mcv@...> wrote:
>
>[JER:]
>>> We find loc. *é-i
>[MCV:]
>> Where do we find that?
>
>In IE *é-i 'then, if', Gk. ei etc.
OK, that's *esmi before *-sm- was inserted.
>The o-stem loc. is oíkoi, but
>that can be a levelled form, which the dialectal variant oikei
>cannot. That presupposes a pronominal or adverbial survival of *(-)e-
>i in thematic stems.
I have no explanation for <oikei> at the moment, but it
can't be a thematic form.
>> >, and *-en-i of r/nt-stems, both with distinctly word-final
>treatment of the part
>> >preceding the *-i.
>>
>> The treatment in the r/n-stems is distinctly non-word-final:
>> otherwise we would have had *-er-i.
>
>There is room for more than just plus and minus here. The boundary
>before enclitics does not have to operate just like absolute final,
>nor obey word-interior rules. And in fact we know there was a some
>strengthening of the articulation in that place in IE. We need room
>for what we find.
And what we find is that the locative is a weak case, which
means that the ending used to be syllabic, even before the
addition of *-i.
My view on the oblique case endings is very simplistic.
There were only three: genitive, dative/locative, and
ablative/instrumental. In the normal PD type, this gives:
[N. *h2ák^-mo:n, A. *h2ák^-mon-m.]
G. *h2ak^-mán-âs -> *h2(a)k^-mén-os
DL *h2ak^-mán-a -> *h2(a)k^-mén-(i) = loc
IAb *h2ak^-mán-ât -> *h2(a)k^-mén-ot(i) = abl.
The vowel of the G. and I/Ab. desincence (-VC) is lengthened
by the svarita rule, but not the vowel of the DL (-V).
In the HD type, we have:
[N. *p&2té:r, A, *p&2térm.]
G. *pah2-tar-ás -> *p&2trés
DL. *pah2-tar-á -> *p&2tré(i) = dat.
IAb. *pah2-tar-át -> *p&2tréh1 = ins.
The AD type had a heavy root syllable (and therefore no
svarita lengthening [only Szemerényi lengthening in the
nom.] in the strong cases), which caused an accent shift to
the lengthened vowel in the G. and I/Ab.
[N. *póntoh2s, A. *pónth2m.]
G. *pu:nt-áh2-âs -> *p&nt&h2á:s > *pn.th2ós
DL. *pu:nt-áh2-a -> *pn.táh2i = loc.
IAb. *pu:nt-áh2-ât -> *p&nt&h2á:t > *pn.th2ót = abl.
When later the need arose to create a dative and an
instrumental for the PD/AD types, or a locative for the
AD/HD types, these forms were built analogically:
dat. *h2(a)k^-men-éi
ins. *h2(a)k^-men-éh1 or *h2(a)k^-(m)n-éh1
dat. *pn.th2ái
ins. *pn.th2áh1
loc. *p&2téri
This last, analogical, form is the only one where the
locative pretends to be a strong case.
As to the "water" word, there seem to be several variants
around. Hittite wa:tar, wetenas seems to point to a stative
paradigm:
wá:dan -> wódr.
wa:dána:s -> wédnos
wa:dána -> wédn(i),
with a collective plural:
wadá:n(h2) -> udó:r
Declined oblique forms:
wada:nás -> udéns
wada:nái -> udén(i).
On the other hand, Vedic seems to point to a slightly
different (AD) paradigm:
ú:dan -> wódr.
u:dána:s -> udnós
u:dána -> udén(i)
(it makes no difference for the collective:
udá:n(h2) -> udó:r
uda:nás -> udéns
uda:nái -> udén(i)).
=======================
Miguel Carrasquer Vidal
mcv@...