Deriving Syntactic Relations PDF Download

This dissertation places case, agreement and Voice phenomena in syntax. It argues that the derivation is driven by so-called derivational features, that is, structure-building features (Merge) and probe features (Agree) (Heck and Müller 2007 and Müller 2010; see also Chomsky 2000, 2001). Both types are essential in deriving case and agreement in the clausal domain and DP-internally. Feature values assigned by Merge take effect immediately whereas feature values assigned via Agree take effect at Spell-Out. This has the effect that Merge can overwrite Agree relations. I argue for a clear boundary between the syntactic and the morphological component regarding how case is assigned and agreement derived, placing Agree, Merge and case assignment in syntax whereas a translation of case assignment into morphological case and agreement takes place in the morphological component. Case morphology is the result of a three-step process: (i) A syntactic relationship with a functional head (e.g., Agree with Voice); (ii) a morphological translation of that relationship into a case feature (e.g., from syntactic STR to morphological ACC); and (iii) a morphological realization of that feature at Vocabulary Insertion. I argue that there are three types of syntactic case: structural, inherent and quirky case. Structural nominative case is either the result of structural case assignment or the realization of unassigned case. If a DP has not been assigned case by Spell-Out, its syntactic case is determined as STR. Structural and quirky case is often assigned by Voice via Agree but inherent case is assigned by Appl via Merge. Furthermore, the dissertation studies the interaction of Voice, case and implicit arguments. It provides new analyses for various constructions in Icelandic where the dichotomy between active and passive breaks down. As I demonstrate, passive and active are labels for a collection of properties of VoiceP, where these properties may vary partially independently, yielding constructions that do not fit the traditional labels. I refine and improve our understanding of the nature of implicit arguments and how they interact with different Voice types. Following Landau's (2010) distinction between Weak and Strong Implicit Arguments (WIA and SIA), I extend Legate's (2014) analysis of the New Impersonal Construction to other constructions. I propose that WIAs are not always projected but when they are, they bear case. Furthermore, I propose that weak implicit arguments have an overtly realized counterpart, which I call Weak Explicit Arguments.