Arguments for successive-cyclic movement through SpecCP

Successive-cyclic A'–Cmovement derivations exploiting SpecCP as an intermediate landing-site deserve careful scrutiny. As a companion to Den Dikken’s (2009a) case for a typology of A'–Cdependencies that includes successive-cyclic movement via vP–edges, resumptive prolepsis, and scope marking, but not successive-cyclic movement via SpecCP, this paper demonstrates that the arguments accumulated in the generative literature in favour of successive-cyclic movement via SpecCP are invalid. To the extent that any of these arguments implicate SpecCP at all, they never make reference to SpecCP as an intermediate stopover point: they are arguments either for terminal movement to a subordinate SpecCP or for successive-cyclic movement via intermediate stopovers in positions other than SpecCP.