On Thu, 3 Jul 2003 10:47:05 +0200, =?iso-8859-1?Q?J=F6rg=20Rhiemeier?=
<[log in to unmask]> wrote:

>I am not an expert on Kartvelian, but I found some information on this
>in Gamkrelidze's and Ivanov's book, _Indo-European and the Indo-Europeans_.
>First, the phonology.  PIE, if one accepts glottalic theory (which I do),
>had a three-way distinction between voiceless, voiced and ejective stops,
>as does Kartvelian (and other Caucasian languages); and apparently,
>Kartvelian has an ablaut system similar to that of IE.

It seems to me that ejective stops can also be called aspirated stops.
I've seen some e/o variations in Georgian morphology, particularly in
affixes, but nothing as seemingly paradigmatic as in PIE.  Can you give any
examples of Kartvelian Ablaut?

Also, regarding Ablaut, my hypothesis is that most of the contrasts in
vowel quality (e vs. o) were originally contrasts in tone.  It is well-
noted that vowels did not undergo alternation in the presence of
a "laryngeal" (?, h, x, etc.).  In certain places, I think that contrasts
in vowel quality were caused by other (unknown?) factors.

>The common points in syntax are SOV order (a frequent pattern that
>proves nothing and is also shared by Proto-Uralic) and active-stative
>argument marking.  That is, intransitive subjects are marked like
>transitive subjects if they are agents (as in 'The man runs'),
>but like transitive objects if they are not (as in 'The stone falls').
>Typically, only animate nouns can take agent marking.

SOV is, apparently, a very ancient (and likely original) word order.  In my
opinion, the first great split within Nostratic occurred when the southern
speakers (> Afro-Asiatic) adopted VSO instead of SOV.

I've seen people try to prove that Proto-Uralic also had stative verbs, but
I simply see no evidence for it except for the Hungarian indefinite
conjugation.  Then again, why would the PU stative paradigm disappear
everywhere but Hungarian?

>This pattern is partly preserved in Kartvelian languages, and traces
>of it can be found in IE (active vs. stative verb endings, avoidance
>of inanimate transitive subjects, syncretism of nominative and
>accusative in neuters (animates had an agentive case in *-s
>and an objective case in *-m, later to become nom. and acc.
>respectively, while inanimates had an unmarked objective case
>and no agentive case).

The origin of the animate sigmatic nominative and the nominative-accusative
syncretism in inanimates is of great interest to me.  My opinion is that
the former was originally an ergative marker, which in turn came from the
genitive.  Syntactically/grammatically, this makes sense: the genitive case
is the case of origin, and only animate nouns can "originate action" (i.e.,
perform an action).  The problem lies in reconciling this with the current
reconstruction of PIE.  As for the nominative-accusative syncretism, it is
obvious that inanimate nouns could never be grammatical agents, and thus
never had a true nominative in PIE.  When the rules of grammar shifted in
later times, speakers used the most common case for inanimate nouns -- the
accusative -- as the nominative as well.  Does this make sense?  However,
it is of interest that this only occurred with thematic inanimates;
athematic inanimates never had a marked accusative case.  I think the most
likely reason for this is that the athematic inanimates were an older
class; they were root nouns (which later disappeared/became thematized),
i/u-stems (relatively rare), and s-, r-, and n-stems (which appear to have
been rather productive at some point in PIE).  Even after the original root
nouns were thematized, the remaining stems were still able to be
distinguished solely by their endings (except for s-stems in the nominative
singular, which I think became confused with thematic animates, e.g.
*nebhos, *nebhes- 'cloud' < **nebhes).

>Why does this point to a substratum?  It is a well-known fact that
>learners of a foreign language have the greatest difficulty with
>phonology (leading to accents) and syntax.  To give one example:
>Native German speakers, when speaking English, often use
>the perfect in situations where a native speaker of English
>would use the simple past, but the perfect is used in German.
>(The German perfect has a wider range of meaning than the
>English perfect.)  So when a population switches to another
>language, features of their old language tend to be carried over
>into the new language in the fields of phonology and syntax
>rather than morphology.

There could very well have been one or more substrata in PIE; the problem
lies in finding it.

>Yes.  And possibly, transitive verbs agreed with both arguments,
>as would be the typical pattern in an active-stative language.
>The thematic vowel might be the last remnant of a 3rd person object marker.

I have also considered this.  However, there does not seem to be a clear
pattern whereby transitive verbs are always thematic, and intransitive ones
are always athematic.  For example, *gwhen- 'strike, slay, kill' is
athematic (*gwhenmi 'I kill') but transitive.  In any case, however, there
is a discernable pattern in the e/o alternations of the thematic vowel: the
o alternation always occurs before a nasal.  Whether this apparent sound-
change was conditional or not remains to be seen.  I think that either
thematic verbs resulted from a 3rd person object marker, or that they were
simply vocalic verb stems (i.e., ending in a true vowel, from the
standpoint of PIE).  I certainly think that the latter is the reason for
the contrast between thematic and athematic nouns.

- Rob