Abstract :

This study aims at experimentally exploring whether reconstruction can actually be taken as a diagnostic of movement in JA. Three experiments were conducted. Experiment 1, an acceptability judgment task, investigated the availability of gap and resumption strategies in JA relative clause and wh-interrogative. Experiment 2, combining both forced-choice and acceptability judgment tasks, tested the availability of reconstruction effects in JA relative clauses regarding the type of the binding principles (A, C, and BVA), resumptive pronouns (Weak and Strong RPs), and islands (strong and weak islands). Adopting the same design of Experiment 2, Experiment 3 tested the availability of reconstruction effects in JA wh-interrogatives. The findings reveal that only resumption strategy is available in JA relative clauses. Furthermore, JA relative clauses do not exhibit sensitivity to islands; JA wh-interrogatives do. Reconstruction effects are available in JA relative clauses irrespective of the type of the binding principle, the type of the RP, and the type of the island, whereas the presence of strong islands blocks the availability of reconstruction effects in JA wh-interrogatives. The theoretical implications of these findings unravel that relative clauses in JA are derived via base-generation rather than A’-movement whereby the referentiality of the RP is achieved by binding.