LINGUIST List 6.636

Mon 01 May 1995

Sum: Null Objects

Editor for this issue: <>

Directory

1. Yan Huang, Re: Summary: null objects

Content-Length: 3022

Some time ago I posted a query about unidentified/non-agreeing null
objects. My question was whether or not there are languages other than
Chinese, Japanese and Korean, which have non-agreeing null objects which
can fit simultaneously in with more than one type of empty category in
the Chomskyan sense, that is, they can have more than one interpretation.
I received three replies, which were from

Charatdao Intratat (g34citchulkn.chula.ac.th)
Shin Ja Hwang (shinja_hwangsil.org)
F.K.L. Chit Hlaing (flehmanux1.cso.uiuc.edu)

In addition David Lee sent me some comments on the original Chinese
example. I'd like to thank them for responding to my query. (I have also
reponded to them individually.)

 Charatdao Intratat pointed out that Thai is another language which has the
quite similar phenomenon to Chinese. In Thai, the null object can have
more than one reading.
 Shin Ja Hwang confirmed that the null object in Korean behaves in a way
that is parallel to that in Chinese.
 Chit Hlaing sent the followin comments:
 -----------------------------------------------------------------------------
 Many Tibeto-Burman and some Tai languages are like this. They allow, even
prefer, a null pronominal whenever the pronominal argument (or
non-argument, for the matter of that) is not stressed/focussed or whatever.
This is the case whether or not there are agreement markers for one or more
of the argument positions.The reason seems to be that, if there are
agrement clitics, as in, say, Chin languages, or Himalayan ones (the
so-called pronominalising languages), these are not morphologically linked
to INFL; either, in such cases, the AGR categories are dominant over INFL
or whatever, but in any cased they are, for instance, left of the verb
while any INFL material is right of the verb, and so on. These are the
F(ree) E(mpty) C(ategory) languages: the parameter setting appears to be
unlikke that for Pro-drop languages, where it is the presence of at least
some agreement-in-scope-of-INFL that triggers the rule: no drop unless the
agreement is rich enough.
 --------------------------------------------------------------------------
 I found all the replies very useful and I would be interested to receive
further comments and information on this topic.