This paper explores a less discussed aspect of head movement by examining two constructions, namely, Across-the-board (ATB) movement and Parasitic Gaps. I reveal an asymmetry between the two configurations: ATB head movement of verbs is attested in Cantonese, but Parasitic Gaps for verbal heads are not. I propose that the unavailability of PGs for verbal heads is not due to the head status of the moving elements: they are ruled out because the possible types for operators are independently restricted. The findings suggest (i) that there is no substantial counter evidence for the unity of movement from PG constructions, and (ii) that ATB movement and PG constructions should receive non-uniform treatment.