agreement

[Edith A Moravcsik]

In response to my suggestion that in Hungarian, the verb agrees
with the object in definiteness, Matthew Dryer pointed out that there are
many languages with alternative forms of verbs such that one is used when
some argument is definite and another when that argument is indefinite
(such as focus alternations in Philippine languages, antipassive versus
basic clauses, or existential versus non-existential constructions) - yet,
linguists would call these cases a kind of constructional contrast rather
than instances of definiteness agreement. Matthew then asked if there was
any argument for describing the Hungarian pattern as definiteness
agreement rather than as a constructional contrast.

If we define verb agreement simply as a construction type where the
(non-lexical) form of the verb co-varies with the values of some
feature of its arguments, then there is, it seems to me, no way
to escape calling focus and voice alternations instances of verb
agreement since here the verb changes shape depending on the semantic role
of its focus or topic or subject argument. But, as Matthew pointed
out, this is intuitively wrong: focus and voice alternations seem to be
somehow different from verb agreement. The question is how we could
explicate the difference.

Perhaps the way to go is to restrict verb agreement to cases where
the agreement feature is non-relational. Thus, number, gender, person
would be non-relational features but "agent of", "patient of" etc. are
relational features and verb form alternations reflecting these features
of the verb's arguments would not be lumped together with the former.
The former patterns would be called agreement and the latter, something
else such as constructional constrast?.

If so, then what about definiteness? If definiteness is a non-relational
feature, then cases where verbs change their form depending on
the definiteness of their argument count as agreement. If
definiteness is a relational feature, then there is no such thing
as definiteness agreement.

It seems to me that the former is correct: definiteness seems more
like number and gender and thus not a relational feature.

If therefore we admit definiteness agreement as a possible phenomenon,
this would still exclude the focus constructions in Philippine languages
from the domain of agreement - or at least those in Tagalog. In Tagalog
the verb changes affixes depending on which semantic argument is chosen to
be the "ang-phrase". But, although "ang-phrases" are said to be generally
definite, they are not always definite nor are all definite NP-s marked
with "ang". Instead, Schachter and Otanes call them topics - a
notion that overlaps with definite but is not the same. (Schachter and
Otanes 1972, 95-97).

What do yoy think?

Edith



   ************************************************************************
			 Edith A. Moravcsik
			 Department of Foreign Languages and Linguistics
			 University of Wisconsin-Milwaukee
		         Milwaukee, WI 53201-0413
                         USA

			 E-mail: edith at uwm.edu
		         Telephone: (414) 229-6794 /office/
				    (414) 332-0141 /home/
		         Fax: (414) 229-2741