LINGUIST List 3.72

Fri 24 Jan 1992

Disc: Nominative in Non-Finite Clauses

Editor for this issue: <>

Directory

1. , Nominative in Non-finite Clauses

As Jonathan Bobaljik rightly observes, ergative languages which allow only
absolutive arguments in non-finite clauses (such as, apparently, Eskimo-Aleut)
are not a problem for traditional GB theory, at least given certain assumptions
about absolutive and ergative case (specifically, that ergative case, but not
absolutive case, is assigned VP-externally).
 I apologize for not having pointed out that Lezgian shows ergative/absolutive
(or ergative/nominative) case marking and that the Lezgian "nominative" is an
absolutive case (in fact, I call it Absolutive case in my forthcoming grammar).
The reason for this omission is that it doesn't matter (and here comes the
surprise for Jonathan): Lezgian non-finite clauses show the same case-marking
as finite clauses, including ergative arguments. For example, (1) is just
as good as (2).

(1) Didedi-z gadadi ktab qacu-na k'an-zawa.
 mother-DAT boy(ERG) book(ABS) buy-NONFIN want-PRES
 'Mother wants the boy to buy a book.'

(2) Didedi-z ktab stold-a xa-na k'an-zawa.
 mother-DAT book(ABS) table-INESS be-NONFIN want-PRES
 'Mother wants the book to be on the table.'

Thus, Lezgian behaves just like Dravidian, and the accusative vs. ergative
distinction is irrelevant here. It seems that the Lezgian facts cannot be
simply explained away and the problem for GB remains.

Martin Haspelmath, Free University of Berlin