The nature of clause-internal Ā-movement of subjects has been a topic of continuous debate, since the presence of such movement is often hard to diagnose clearly. Additionally, while parasitic gaps are a well-studied phenomenon, the interaction of parasitic gaps and subjects has rarely been explored in detail. In this presentation, I argue that the behavior of subject parasitic gaps provides new evidence for a theory in which (at least in the constructions under consideration) clause internal subject Ā-movement is impossible. I situate this finding in the context of the hypothesis of anti-locality, which posits that certain movements fail due to being too short. I show that anti-locality analyses of phenomena like the that-trace effect extend naturally to the PG facts I focus on, though several puzzles will remain.
