The nanosyntax of case

Abstract

This dissertation proposes a new approach to case. It unifies its syntax, morphology and semantics in a simple, fine-grained and restrictive picture. One of the assumptions frequently made in works on case is that cases such as nominative and accusative are not primitive entities, but they are each composed of various features. The central hypothesis of this dissertation is that these features are universal, and each of them is its own terminal node in the syntactic tree. Individual cases thus correspond to phrasal constituents built out of these terminals. The idea that syntactic trees are built by Merge from individual atomic features is one of the core principles of a cartographic approach to syntax pursued by M. Starke: Nanosyntax. Hence “The nanosyntax of case.” I motivate the approach on the material of case syncretism. I propose a hypothesis according to which case syncretism across various languages obeys a single restrictive template. The template corresponds to a cross-linguistically fixed sequence of cases, in which only adjacent cases show syncretism. In order to derive this, I argue that case features are syntactic heads, ordered in a universal functional sequence. If this is so, it follows that these sub-morphemic features interact with core syntactic processes, such as movement. The prediction is borne out: the interaction of (phrasal) movement and the fine-grained syntactic representation derives a typological generalization concerning cross-linguistic variation in the amount of case marking (Blake’s hierarchy). Additional facts fall out from the picture: the role of functional prepositions, prepositional syncretism, case compounding, and preposition stacking. I further investigate in detail the spell out of these highly articulate structures. I follow Starke (2005) and propose that individual morphemes spell out phrasal constituents of varying size, and that their insertion is governed by the Superset Principle. I argue that phrasal spell out is both empirically required, and theoretically beneficial: it simplifies the overall architecture of grammar. In particular, there is no part left to play for a separate morphological structure. With the proposal in place, I observe that there are generalizations which connect the proposed representation and the DP external syntax. To account for this, I adopt the Peeling theory of movement (Starke 2005). The theory says that arguments are base-generated with a number of case projections on top of them, and they strand these projections when they move up in the tree. The theory is shown to capture the initial observations, as well as additional generalizations: Burzio’s generalization among them. The resulting theory does not introduce any domain specific tools to account for case: its representation corresponds to a binary syntactic structure, its computation corresponds to syntactic movement.