Syntax; Generative typology

I am a 4th year PhD student, with a primary interest in locality, phase theory, agreement, case, ergativity, and Austronesian voice systems, all from a comparative perspective. I investigate the viability of a narrow-syntactic system where operations are extremely local, basically mediated by (Re)Merge under sisterhood. 

I have insofar worked on three representative case-studies of such a Merge-based system. The first is Long-Distance Agreement (in Indo-Aryan, Chukotko-Kamchtakan, Nakh-Dagestanian, Algonquian). I take LDA to be mediated by (possibly covert) remerge with the verbal target either of the DP controller itself, or of the complement clause that contains the DP controller at its edge, via clausal pied-piping plus transitive (spec-head) agreement. The main empirical arguments stem from the movement properties and constraints that LDA is sensitive to, and the direct effect it has on scope and information-structure. Thus, LDA reduces either to 'semantic' agreement that signals LF-related properties of scope-discourse, or to the morphological reflex of independent narrow-syntactic operations (such as successive-cyclic movement, (hyper)raising, scope-taking, case-licensing, scrambling, etc.). This is not to deny the existence of agreement from a distance, without movement (including LDA in Greek, Romance and some constructions of Algonquian); but given its dissociation from any syntactic or semantic consequences, I take it to be a(n early) post-syntactic process. This project has implications on the modularity of grammar, ultimately speaking in favor of agreement as a post-syntactic phenomenon (Bobaljik 2008).

The second case-study is Phase Unlocking (Rackowski and Richards 2005): the idea that extraction from within the domain of a phase requires prior Agree of the phase with a higher probe, which can then attract a phase-embedded goal, ignoring the intervening phase by the Principle of Minimal Compliance (Riachards 1998). With novel evidence from Austronesian Malagasy and Tagalog, I argue that 'unlocking' of a phase is contingent on movement of the phase to the edge of the probing head H, before a phase-embedded goal can subextract to an outer specifier of HP. The contingence of 'unlocking' on remerge can tackle several open questions, such as the complementarity of unlocking with successive-cyclic movement, its compatibility with spell-out, or how to structurally derive Minimal Compliance. Importantly, it reduces 'unlocking' to a broader strategy of obviating intervention; i.e., remerge of the intervener. This strikingly resembles various instances where extraction out of a strong island is exceptionally possible under pied-piping of the entire island. Thus, Agree is insufficient to render a locality domain transparent, while remerge must be both necessary and sufficient. Ultimately, I aim to argue in favor of the reality of phases as absolute cyclic locality domains, not just as relativized interveners (pace Halpert and Zeiljstra 2026).

The third case-study has its roots in an underdiscussed idea of Sportiche (2005): that nominal arguments are borne as bare NPs at their base θ-position and become DPs by moving to dedicated functional projections, where they incrementally acquire their nominal layers. This may have far-reaching consequences: it derives the co-variation of a nominal's structural size with its structural height and semantic interpretation (basically, the effects of Diesing 1992 et seq.); it reconciles the apparent conflict between the need for lexical V to select for the N (rather than the D) head (Bruening 2020) and the evidence that the head of the nominal is ultimately D (rather than N); it may accout for the definiteness effect (Milsark 1977) and the ban on quantifier stranding in θ-positions; and it allows to reconcile syntactic nominal licensing with a (post-syntactic) Dependent Case alogorithm that yields different alignment systems while dispensing with Agree-based case-assignment by functional heads.

Moreover, me and Magdalena Lohninger (University of Salzburg) have a joint project on Austronesian pivots and voice-systems, focusing on the (non-) subject status of pivots, and on the status of voice as a discourse-configurational or passive system. With a core sample of 6 languages (Acehnese, Balinese, Cebuano, Malagasy, Standard Indonesian, Tagalog),  we argue that pivothood is not a uniform phenomenon: in some languages (Cebuano, Tagalog), pivots are V2-like topics, and the voice-system is dissociated from case-licensing and grammatical roles; in some languages (Acehnese, Indonesian), pivots are derived subjects, and voice-system is a non-canonical passive feeding case-licensing, free of information-structural restrictions; and in some languages (Balinese, Malagasy), pivots are in-between hybrids, both V2-like topics and surface subjects. Our primary diagnostic comes from obligatory control; i.e., whether a language controls the embedded pivot, regardless of θ-role, or the embedded external argument, regardless of pivothood. The choice of controllee systematically correlates with whether a language can only license overt non-pivot external arguments via strict adjacency to the verb or not, which we tie to the interaction between pivot promotion and case-assignment. Ultimately, we aim to establish that 'surface subjecthood' (in particular, subject properties, such as control and anaphor binding) is structurally determined at the highest A-position within the argument/event domain, as delineated by VoiceP.  

Other topics I have worked on include syntactic ergativity (my MA thesis), Clitic Left Dislocation, implicit contextual variables and domain restriction, the Plural Pronoun Construction (with a focus on Greek), and the effect of de re/de se readings on locality boundaries for cross-clausal operations.