Resumption in focus(-raising)

Hungarian focus-raising constructions had traditionally been analyzed as instances of long-distance movement (É. Kiss, 1987; Marácz, 1987; Horvath, 1995, 1998; Lipták, 1998). Recently, new empirical evidence (Gervain, 2005) has shown that while some speakers do indeed derive the construction by movement, others use a resumptive strategy instead. The current paper investigates the exact nature of this resumptive dependency, asking whether the resumptive is pronominal or variable-like. Using islands, strong crossover effects, reconstruction and parasitic gap licensing as diagnostic tests, it is shown that resumptives in Hungarian FR are derivationally not traces, since they do not result from movement, but they are variables, requiring local A’-binding.