LFG Questions

[Brian O'Herin]

I've been reading about LFG and hanging around LFG events for several years
now, and I figure it's about time for me to start trying to write some
within the theory. I have worked in Abaza, a Northwest Caucasian language,
and there are some constructions in the language I would like to
investigate. I'm wondering if anyone has looked at these issues before or is
maybe interested in working together on any of these topics. Even getting
pointed in the right direction would be helpful.

 

First, I'm interested in the difference between predicate and attributive
adjectives. I recently came across the PredLink idea in the "Grammar
Cookbook" and this was new to me. Is it still being used? I had a student a
couple years ago working on a Sino-Tibetan language with no overt copula,
and we spent a long time trying to figure out how to deal with that. We got
hung up on the fact that we needed two lexical entries for almost every
adjective - one with a SUBJ and one without. (Not to mention the problems
that causes when one comes to predicate nominal constructions.) I think I
can see how the PredLink might solve that, but it seems to me it would need
to be introduced by annotation on a c-str rule rather than associated with a
lexical item like a copula. Abaza is different: there is no copula (actually
there is a copular construction, but I'm pretty sure it's borrowed, it's
used more for predicate nominals than for predicate adjectives, and it isn't
nearly as widely used as the non-copular construction), but there is
predicative morphology on the predicate adjective, indicating tense and mood
(in the stative system, as opposed to dynamic, which is what most verbs
are). In that case, it seems the PredLink would be introduced by an
annotation on the lexical entry for the adjective (specifically introduced
by the morphology). Prior to coming across the PredLink analysis, I had been
assuming that the predicative morphology would introduce a SUBJ, but there
are complications with that, too.

 

(Incidentally, I'm also quite interested in the interaction between
morphology and syntax within LFG. Abaza has LOTS of verb forms, with a  very
productive morphological component. In some ways, it would be a lot easier
if there were no Lexical Integrity Principle, since then I could deal with
the morphology in the syntax. That's a whole separate issue.)

 

The second issue I'm especially interested in has to do with a construction
I've not come across in any other language, at least not exactly like it.
Abaza has overt agreement with just about everything - subject, object,
indirect object, applied object(s), causative subject, possessor,
postpositional objects, etc. The pattern is ergative-absolutive, and it
distinguishes person, number, gender and animacy. In addition, there is a
"wh-" form (for content questions and relativized positions), which
overrides any other features. That much is not all that difficult to deal
with. The challenge is that if there is an NP elsewhere in the sentence that
is coreferent with the element that is marked by wh-agreement, the agreement
for that other NP also shows up as wh, even if it's not in a wh position
itself. The main disclaimer is that it has to be in the right structural
configuration. In my dissertation using GB, I analyzed it as c-command, but
I'm not 100% convinced that's right. I'm trying to go through texts to find
as many cases of this as I can, but it's not especially common. In terms of
analyzing it within LFG, I'm not sure even where to begin, so I would
appreciate any help or direction I can get.

 

Brian

 

Brian O'Herin, PhD

International Linguistics Coordinator

SIL International

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lfg/attachments/20101025/385661b6/attachment.htm>