Cross-clausal movement and its limits

Constraints on movement are of central concern to syntactic theory. Recent work has focused on the relationship between the height of the landing site and locality restrictions. Keine (2020) argues that the higher the landing site the fewer locality restrictions are imposed on the movement (“higher is freer”). In contrast Deal (2017) argues that a lower landing site can lead to movement being less restricted due to Delayed Opacity, i.e. “lower is freer”. In this paper we argue that, in Kalaallisut (Inuit, Greenland), it is not the height of the landing site that determines whether movement across a finite clause boundary is possible, rather it is the nature of the features driving the movement. In particular, hyperraising is driven by a composite ϕ+D probe, which means that the embedded CP does not act as an intervener for Agree with a goal inside the embedded clause (Halpert 2019). In focus fronting, on the other hand, the embedded CP does intervene for Agree with a focused element inside the embedded clause, which accounts for the clause-bounded nature of focus fronting. In addition to deriving the differing locality profiles of hyperraising and focus fronting, the analysis of focus fronting generalizes to wh-movement and also accounts for apparent long-distance A-bar movement. The final segment of the paper lays out a broader cross-linguistic typology of locality profiles for A and A-bar movement and suggests that it arises from variation in probe specifications and in the featural content of C.