118 Julia Horvath
of distribution involving indeclinable nominals.^1 I propose below an analysis for the
source of the relevant pattern that resolves the puzzle, rendering it a direct consequence
of an independently attested property – a Case Realization Requirement – of oblique
case in the language, in conjunction with a particular new conception of the nature
and assigner of oblique case in general. Specifically, we will be led to the conclusion
that contrary to appearances, and in contrast to the standard assumption, oblique (lexi-
cally governed) cases can only be assigned by (overt or null) instances of the category
P. Verbs or other lexical categories do not assign oblique cases to their complements.
The facts analyzed will provide striking support for the innovative approach to
morphological case advanced in Pesetsky (2013), which claims that the various tra-
ditionally recognized grammatical cases are not sui generis features in the theory, but
can be reduced to affixed copies of particular part-of-speech categories. More spe-
cifically, based on a detailed examination of the complex and peculiar morphological
case properties of Russian nominals, Pesetsky (2013) puts forward the hypothesis that
rather than having traditionally stipulated case categories of their own, morphologi-
cal nominative reduces to a realization of the category D, accusative to the category V,
genitive to N, and oblique to P. If it turns out to be empirically tenable, this approach
would clearly be a conceptually highly desirable move.
The Serbian/Croatian phenomenon discussed below will be shown to provide
empirical support for the specific claim that the so-called “oblique” cases, apparently
assigned to complements by various lexical heads, are to be analyzed as affixal copies of
the part-of-speech category P, assigned uniformly by an overt or null P head.^21.^ Locative is also an oblique case in Serbian/Croatian, however, unlike the dative and the
instrumental, it is never governed by V or N, only by P; thus it does not manifest the puzzle
posed by the latter two cases to be discussed here. Genitive case is also often, though not
uniformly, considered an oblique case, however, it involves additional complexities (see e.g.
Franks 2002; Bošković 2010). Thus, the examination of its behavior with regard to the distri-
butional puzzle to be discussed in the sections below is beyond the scope of the present note.
2. The point of the present note is the presentation of novel evidence for the reduction
of oblique morphological cases to the part-of-speech category P, as expected based on
Pesetsky (2013). The general idea of some intrinsic relation between adpositions (P) and non-
structural case (often referred to as K) has appeared in the literature in various forms; it has
been suggested specifically for Slavic on different grounds, among others, in Bošković (2006,
2010 ). Some evidence in support of the existence of extra structure in oblique (inherent case
bearing) noun phrases from extractions and phases (in conjunction with an assumption of
anti-locality) is presented in Bosković (2010). For a discussion of the nature and status of
“oblique” case in relation to the structural/inherent dichotomy of the GB case theory, see
Franks (2002). A proposal for the different morphological cases of Slavic to be identified with
particular functional categories is made by Bailyn (2004).