Miguel:
>The construction you give above is redundant in that it has
>both an interrogative *kW- and a deictic *h1e- (Bob what? Jim
>wrong being. That thinks.).
Not totally redundant. I was thinking of Mandarin at the time
I came up with this idea. *?e is being used to seperate the
main verb from the verb of the subordinate phrase and might
be translated as "thus", while the *kW-particle signifying "who"
or "which" marks the beginning of the clause after the
specified subject. So the two particles are marking both the
beginning and end of the clause, which is partly "redundant"
but partly not because it enables better comprehension.
The issue in Mandarin, an SVO language, that inspired me was from
a conversation with a Chinese friend I know. I asked him advice on
how to say similar things in his native language. In Mandarin, the
particle /de/ is used to mark the end of the clause and this clause
is placed BEFORE the noun it modifies if a noun is specified (eg.
/Qu-le zai Beijing de ren/, "Went to Beijing who person" or
"The person who went to Beijing."). (In actuality /de/ is also used
as a genitive marker in /Bob de/ "Bob's" and can be thought of
in general as a multi-purpose particle that simply links the
preceding modified to the following modifier.)
Now, I notice my friend may place /ge/ (a noun classifier normally
used with numerals like /yi-ge xiangjiao/ "one banana") before a
clause and this nicely seperates the main verb from the following
verb of the clause, if I do say (eg: /Wo shi ge qu zai Beijing de
ren/, literally "I am one go to Beijing who person", and meaning
"I am the person who is going to Beijing"). Here, /ge qu zai
Beijing de/ is the clause with both /ge/ "one" and /de/ "who"
marking the beginning and end of it. Again, perhaps redundant as
you say but it really improves comprehension compared to a sentence
like /Wo shi qu zai Beijing de ren/ "I am go to Beijing who person/
where the two verbs are said in succession without warning,
requiring extra brain processing.
So, Mandarin justifies the pattern I observe for an earlier stage
of IE and helps to elaborate on a possible transitional stage
between a language with the *kW-particle AFTER its clause and a
language with the same particle BEFORE the clause as we find in
PIE itself. The intermediary then is a language where both the
beginning and the end of a clause is explicitly marked with _two_
particles.
>Returning to relative clauses, a very common type is that of the
>genitive construction, as presumably in Basque,
>[...]
>emakumeari liburua eman dion gizona
>woman-the-DAT book-the given has-he-it-to her-of man-the
>The man of he having given it to her the book to the woman
>The man who gave the book to the woman
Yes, and the same concept exists in Mandarin (but in a different
order) as I've said. The equivalent in Mandarin would be
/Bei-le nuderen yige shu de nanderen/ "Gave woman one book who man"
or in another way "Gave-woman-one-book's man".
However this would not be the case for pre-IE because there was
already a *kW-particle that served the purpose for clauses and
a seperate ending in *-s/*-os for genitives. I think in pre-IE
there is a greater link to be found between noun or adjective
stems in later *m or *q and ancient clause formations. In
other words, these stems are fossilized clauses, but I digress
now.
>Turkish has a similar construction, except that the genitive is
>expressed using a possessive pronoun (cf. Dutch Jans boek / Jan z'n
>boek: John's book, John his book).
Reminds me of Burushaski.
>The interesting thing about the Turkish construction [...] It is
>easy to imagine scenarios where constructions such as this gave
>rise to either the PIE nominative *-s or the subjunctive in *-�
>or both.
Actually I'm sticking with the more trivial solution: Definite *so
as the source of the nominative. It should be also noted that the
genitive (or if available, a special partitive) can be used to
convey indefiniteness (French /Je bois de l'eau/ using the genitive
particle /de/ "of, from" for a partitive marker, the equivalent of
Finnish's /Vetta" juon/ that uses a specific partitive case, all
to convey "I drink some water" or rather "I drink from the water"
as we may say in English).
I think that Old IE once had a partitive case (the ancestor of
ablative *-od) and employed it for the indefinite nominative and
accusative. The definite nominative was unmarked and the definite
accusative (for animates only) was marked in *-m. In addition to
the mix, it used the *s-genitive for possessive constructions and
for the agentive in passive constructions with an unmarked patient.
So to sum up:
animate inanimate
------------------------------------------------
nominative ZERO --
accusative *-m ZERO
genitive *-ese
partitive *-eta
agent (genitive) *-ese
patient (nominative) ZERO
>In fact the PIE triple marking of subject (*-s or *-0) and object
>(*-m or *-0) almost certainly does derive from the amalgamation of
>different verbal constructions using different markers for
>ergative/absolutive and/or nominative/accusative and/or
>possessor/possessed.
Actually, sorta. We may simply conclude that there was an early
grammatical and syntactic distinction between animate and inanimate
nouns in the accusative with an originally unmarked nominative.
Hittite disallows inanimate subjects (which inheirantly makes sense
afterall since inanimates aren't supposed to "move") and may be
reconstructed as a feature of PIE, as well as pre-IE to explain
the lack of *-m for inanimates which don't need explicit marking
since they are _always_ objects.
I think that you are too obsessed with having an immediately
preceding ergative Kartvelian-like stage for IE whereas my
conclusion is that there were intermediary stages in between.
No stage of IE as I define it (between 7000 and 4000 BCE) was ever
"purely ergative".
- gLeN
_________________________________________________________________
MSN 8 with e-mail virus protection service: 2 months FREE*
http://join.msn.com/?page=features/virus