On Mon, 22 Mar 1999, Boudewijn Rempt wrote:

> It must depend heavily on the language group then - my observation certainly
> holds true for the Himalayan languages I have studied - but then I don't know
> much about Australian or Caucasian languages. I'm fairly sure
> about Nepali too (and that would imply Hindi too), and Tibetan, but I'm
> handicapped
> in not having my references here. I'll take a look in Dixon's Ergativity
> tonight.

It's me - Boudewijn Rempt posting from my wife's account at home -
and, to make things worse, follow-upping on myself.

But I've consulted Dixon, and the relevant quote is "... the case
which includes S is 'generally' the unmarked one. It seems that
absolutive is always unmarked with respect to ergative." (Dixon 1994,
58) Also footnote 22: "One is tempted to take the generalisation 'the
case which covers S is most likely unmarked and have zero
realisation', and analogise it to cross-referencing systems,
suggesting 'the bound pronominal paradigm which covers S is more
likely to include (more?) zero realisations than the paradigm which
does not cover S." (Dixon 1994, 68)

I'd be interested in references of examples to the contrary, as Dixon
notes that more research is needed.

Of course, if there's object agreement in the verb, it's likely to
have an overt expression :-), except for the third person, when it's
often expressed by a null-morpheme - or when there's split
ergativity, for instance where first and second person agree
according to an ergative pattern, and third person according to an
ergative pattern, or any other complication. For instance, Ebert
(1994:140) shows both patterns for Phedappe Limbu: agreement with the
Agent only, and agreement with Subject and Object.

In Rempt (1994: 26) I have shown for Gamale Kham an a suffix <-o>,
glossed as 3sA and a suffix <-0>, glossed as 3sS/P.

I've only once made a conlang with shows object agreement, but in
that case I wanted to do it too neatly, and provided different
affixes for S, O and A, a situation that practically doesn't occur

Boudewijn Rempt

[log in to unmask]

(Dixon, R.M.W., 1994, Ergativity, Cambridge Studies in Linguistics
69, Cambridge University Press

Ebert, Karen, 1994, The Structure of Kiranti Languages, Comparative
Grammar and Texts, Arbeiten des Seminars fuer Allgemeine
Sprachwissenschaft Nr. 13, Universitaet Zuerich.)

Rempt, Boudewijn, 1994, The Verbal Agreement System of Four Kham
Languages, Linguistics of the Tibeto-Burman Area, Volume 17.1:1-59.