124 Julia Horvath
However, as pointed out by W&Z (1999), this proposed account is untenable as the
presence/absence of P, e.g. the (overt) preposition sa ‘with’, in fact does not depend on
the need for overt realization of case on the particular noun phrase. On the one hand,
the preposition may fail to be available in spite of the complement noun phrase lack-
ing inflection for case, and on the other, in various cases the preposition appears even
though the noun phrase exhibits case inflection of its own. Thus the distribution of the
Ps under discussion indicates that they are categories selected by the verb (or noun),
rather than being merely default morphological realizations of oblique cases assigned
by the latter lexical heads. On these grounds, the P-as-case-inflection proposal can
reasonably be discarded as a solution to the above puzzle (W&Z 1999, 2001 ).
But now we are back to the original problem. How can we explain the ban on noun
phrases lacking overt case realization in V-governed and N-governed oblique (dative/
instrumental) case positions, while permitting the same noun phrases to appear in
P-governed oblique case positions?- Oblique case is uniformly P-governed: The P-copying proposal
Having concluded with W&Z, and contra Franks (1995), that the Ps occurring in the
cases under discussion are indeed categories taking the noun phrase as their comple-
ment, and are assigners of oblique case, rather than simply case-markers on them, let
us take a second look at the facts that constitute the core of our oblique case puzzle
with a fresh perspective.
To begin with, consider specific approaches to case features and the relevant
assignment/checking process advanced in versions of the theory. If the mechanism of
“case assignment” (whether conceived as actual transfer of the feature or as its check-
ing) involves the same case feature being shared by assigner and assignee, then the fact
that an oblique case assigning overt category (such as P in (11)–(12)) is able to satisfy
the requirement for the overt realization of case should come as less of a surprise. The
examples with a P-governed oblique case as above might in fact be expected not to
need any further realization of case within the noun phrase complement itself. Specifi-
cally, it may be expected that no matter whether the noun phrase lacks inflection for
case, the result will still be grammatical as the requirement of (morphological) case
realization will in fact be satisfied within the case-assignment domain, namely, by the
case-feature of the overt assigner itself, i.e. the overt case-bearing head P.
This initial idea for an account becomes more precise, and in fact can yield the
observed facts straightforwardly, if one identifies “oblique case” as being the categorial
feature P itself, as proposed in Pesetsky’s (2013) theory of (morphological) case.
Based on the analysis of Russian case phenomena, Pesetsky advances an approach
to morphological case that seeks to eliminate the set of traditionally recognized