4 Alexander Grosu
prepositions. Horvath offers a solution to this puzzle that is based on: (i) A generalized
version of Wechsler and Zlatić (2001) case realization requirement, which says that
oblique cases must be overtly realized on some element of the ‘assignment domain’,
the latter taken to consist of the assigner and the assignee; (ii) a proposal of Pesetsky
(2012) to the effect that morphological cases are merely copies of specific syntactic
categories (in particular, Nom, Acc, Gen, and Dat/Inst/Loc are copies of the categorial
features of D, V, N and P respectively), so that ‘assignment’ is reinterpreted as ‘copying.’
(i) and (ii) are shown to jointly account for [A] and [B]. Concerning [B], the overt P,
being part of the ‘assignment domain’, satisfies itself the case realization requirement
(i) under the category-feature copying view of morphological case (ii). [A] follows
too, as V cannot be an oblique case assigner under (ii), and when V appears to assign
oblique case, there is in fact a null P (selected by V) that does the job, but the latter
cannot all by itself satisfy case realization (i). The resolution of the apparent puzzle is
taken to provide evidence that oblique case is uniquely assigned by P, as well as support
for Pesetsky’s (2012) reduction of morphological cases to copied category-features.
As noted earlier, the ensuing chapters are primarily concerned with issues of inter-
nal DP structure. I will now outline the structural proposals made in individual chap-
ters, starting with the internally simplest types of DP and moving on to constructions
with increasing internal complexity, pointing out the interplay of structure, agreement
and case.Małgorzata Krzek focuses on the null DPs that function as subjects of impersonal
constructions in Polish. Such null DPs, although configurationally atomic, are viewed
as complex in terms of their featural make-up. The author adapts the feature geometry
proposed by Harley and Ritter (2002), which was devised for referential null DPs, in
a way that enables it to handle impersonal pronouns. Pronouns are assumed to enter
the derivation equipped with a variety of interpretable, but un-valued, features, which
get valued in the course of the derivation through binding by various operators found
within the clause structure, along the lines proposed by Sigurðsson (2004, 2009 ),
Frascarelli (2007), and Holmberg (2010a, b).
Krzek provides a battery of syntactic tests in support of the thesis that the imper-
sonal constructions of Polish, although incompatible with overt subjects, nonetheless
possess pro subjects. She discusses two impersonal constructions, each of which exhib-
its a characteristic overt morpheme. One of them exhibits the reflexive pronoun się,
and has properties comparable to those of the impersonal si constructions of Italian
discussed in Cinque (1988); the other construction is characterized by the suffix -no/
to attached to an uninflected verb. The author views pro as belonging to the category
Noun, and thus as a complement of some functional head. Whereas in referential null
pronouns pro is the complement of a null determiner marked with a referential index,
in the się construction, się is the functional head that takes pro as complement; this