--- In
cybalist@yahoogroups.com, "tgpedersen" <tgpedersen@...>
wrote:
>
> I am much impressed, but not quite convinced. One advantage of
> seeing the supposed finite verb (with active endings) is that it
> will explain why *-os is both nominative and genitive and *-om is
> both accusative and genitive; thus: the PIE sentence with an
active
> verb is actually a verbal noun with two genitives (or a genitive
and
> a partitive?). Therefore *-os, the 'subjective genitive' marker,
> becomes a nominative marker, and *-om, the 'objective genitive' or
> partitive marker, becomes an accusative marker when the verbal
noun
> is reinterpreted as a finite verb.
I am quite impressed too, but not buying. I take your second
sentence to mean "One advantage of seeing the supposed finite
verb ... [as a verbal noun] is that ...". I have several objections:
1. The nominative and the genitive singular are not identical; the
genitive morpheme has a vowel, the nominative never does. I also
believe they have two different sibilants.
2. The accusative singular is just *-m, while the genitive you mean
will be the genitive plural in *-oom, if not *-oHom. They are not
exactly identical either.
3. I see no reason for the odd distribution that an old subjective
genitive is supposed to become a nominative singular, and an old
objective genitive is later found as a genitive plural and an
accusative singular.
>
> In order to get there, we must see -t as a verbal noun marker (the
> road between adjectival forms of verbs and verbal nouns is short).
> That should take care of the *-gWHen "-killer" argument (since now
> agent nouns have -Ø, action nouns -t).
Stems in *-t are generally agent nouns, not action nouns. That
actually seems to matter very much for this language.
> -m and -s are then either portmanteau morphemes with the double
> meaning "my V-ing", "thy V-ing" respectively, or were once -t-m
and
> -t-s, respectively.
Vaillant actually suggested (in a BSL paper of 1937) a derivation of
2sg *-s from **-t-t. I used to be frightfully impressed by it, but
today I cannot accept it. Stems in *-t are not passive participles
as he wanted them to be, and -tt does not yield -s in any other
cases we know.
If we look across the fence what we find is a structure of agent
nouns (active participles) accompanied by subject pronouns. At least
that is so in Uralic and in Eskimo-Aleut. The object case has an *-m
in IE and Uralic. In Eskimo-Aleut the case-marker *-m is genitive.
That could be its old function. The object function would then be
due to precisely the syntactic reinterpretation you suggest, the
passage from objective genitive to accusative, only the objective
genitive would mark the object of an agent noun, not of an action
noun. That step was not taken by Eskimo-Aleut where *-m is still
genitive.
Jens