name:  
    Adam Przepiorkowski
title: 
    Predicative Case Agreement with Quantifier Phrases in Polish
affiliation: 
    Ohio State University and Polish Academy of Sciences



Predicative Case Agreement with Quantifier Phrases in Polish

In Polish, _predicative_ APs usually agree in case with the NP they
are predicated of; in (1), the subject NP and the predicative AP are
nominative, while in (2), the QP and the predicate are accusative.
(2) reflects the fact that, in Polish, QPs in subject position are
accusative (Franks 1995, Przepiorkowski 1996) and that they are headed
by the quantifier (Saloni and Swidzinski 1985, Przepiorkowski 1996).

What is unexpected and unanalyzed so far is the possibility of the
adjectival predicate agreeing with the genitive NP dependent of the
quantifier, cf. (3).

In order to deal with (1)-(3) we propose to slightly extend the domain
of case agreement: an XP agrees with a phrase YP if and only if either
i) XP agrees with the head of YP, or ii) XP agrees with a specifier of
YP, but only if this specifier bears the same index as the head of YP.
This idea could be implemented in Minimalism by defining the feature
checking configuration to be [specifier, head], as normally assumed,
but also exceptionally [higher specifier, lower specifier] (of the
same head), provided the lower specifier and the head are co-indexed.

We substantiate this idea, first, by arguing that the genitive NP
within a QP is a specifier rather then a complement.  One reason for
this position is that, when a QP occupies an oblique case position
(dative, instrumental, etc.), the NP agrees with the quantifier,
instead of being assigned the genitive case, cf. (4).  As we show in
the full paper, in Polish, such mixed `agreement / non-agreement'
pattern is typical of specifiers.

Second, we show that the index of the QP and, hence, the index of its Q
head is the same as the index of the NP specifier.  This is clear from
the binding facts in (5)-(6).  In (5), the anaphor _siebie_ is plural
and masculine, just like the NP _facetow_.  Although Polish anaphors
do not inflect overtly, they agree with their emphatic modifiers
(_sam-_), which do show number and gender overtly.  Since the index of
the anaphor is plural masculine, so must be the index of the subject
QP.  Similarly, in (6), both the NP and the anaphor (and, hence, also
the QP) are plural feminine.

Finally, we look at constructions similar to QPs in that they involve
a head noun occurring with a genitive AP specifier in direct case
positions, cf. (7), and with an agreeing specifier in oblique case
positions, cf. (8).  Interestingly, these constructions do not allow
case optionality analogous to (2)-(3); here, the predicative modifier
may agree with the head only, cf. (9).  Also NP subjects with
genitive possessives, arguably specifiers, do not show such an
optionality.  This difference between (2)-(3) vs. (9) and NPs with
possessive specifiers is expected on the analysis which links the
optionality of case agreement to the identity of indices: in (2)-(3),
the QP has the same index as its NP specifier, while in (9) the AP
specifier does not have an index at all (because adjectives do not
bear indices) and, in case of genitive NP possessors, the possessor
always bears a different index than the possessed.

We provide a formalization of these ideas within Head-driven Phrase
Structure Grammar and argue against some potential alternatives.


(1)
Janek    jest mily. 
John-NOM is   nice-NOM

(2)
Kilka     drzew     bylo wyrwane  z ziemi. 
a few-ACC trees-GEN was  torn-ACC from earth

(3)
Kilka     drzew     bylo wyrwanych z ziemi. 
a few-ACC trees-GEN was  torn-GEN  from earth

(4)
Pomagalem  pieciu   kobietom. 
helped-1SG five-DAT women-DAT

(5)
Pieciu   facetow  zobaczylo siebie samych       w lustrze. 
five-ACC guys-GEN saw       SELF   EMPH-PL.MASC in mirror

(6)
Piec     kobiet    zobaczylo siebie same        w lustrze. 
five-ACC women-GEN saw       SELF   EMPH-PL.FEM in mirror

(7)
Cos ciekawego przyszlo mi do glowy. 
something-NOM?/ACC? interesting-GEN came me-DAT to head
`Something interesting occurred to me.'

(8)
Przysluchiwalem sie  czemus        ciekawemu. 
listened-1SG to REFL something-DAT interesting-DAT

(9)
Cos               ciekawego       bedzie  mile   widziane / *widzianego. 
something-NOM/ACC interesting-GEN will be nicely seen-ACC/*GEN
`Something interesting will be well-received.'