Re: Lexicalising Ergativity

Hallo!

On Tue, 22 Jun 2004 10:28:43 +0100,
Joe <joe@...> wrote:

> william drewery wrote:
 [...]

> >I may have misunderstood you here, but are you saying
> >that ALL erg/abs languages are in fact split-S
> >systems? Because i thought that that Split-S was less
> >common than one or the other.
> >
> >
>
> I think she's confusing it with something else.  Some languages are
> Split-S(generally Ergative in a Past tense), such as Nepali. This is not
> ever so common, but it occurs in many of the Indic languages.  Georgian,
> too, I believe.  But there are also many 'pure' Ergative languages -
> that is, they are ergative in all tenses.
There are several kinds of splits.  First, as you mentioned it,
there is the type exemplified by Indo-Aryan languages, where ergative
marking occurs in the past tense (or the perfective aspect) while
accusative marking in the present tense (or the imperfective aspect).
Then, there are the split-S and fluid-S systems of the kind I have
demonstrated with my conlang Old Albic, where active verbs mark their
subjects like transitive subjects and stative verbs mark their
subjects like transitive objects.  (See my post earlier in this
thread.)  Georgian is actually accusative in the present
and split-S (rather than purely ergative) in the aorist.
And third, there are languages such as Dyirbal (an aboriginal
Australian language) with ergative marking on nouns and accusative
marking on pronouns.  (See also Thomas Wier's Phaleran for a
conlang example of such marking.)  Of course, one can combine all
of these splits in a conlang (as far as I know, Danny Wier is
planning something like that for Tech, once he learns not to
strange himself with 200-odd phonemes).

>       However, no Ergative language
> is completely devoid of accusative characteristics, nor the other way
> round.
True, at least for natlangs.

Greetings,

Jörg.