Movement and crossover in three languages.

In recent work, Bhatt and Keine (2019, 2023) have shown that Hindi local scrambling obviates Secondary Weak Crossover (SWCO) but not Secondary Strong Crossover (SSCO). The purpose of this paper is two-fold. The first is to show that this property is quite widespread, as it applies in a virtually identical manner to Italian Clitic Left Dislocation (CLLD) and to PP-fronting in English, facts that had gone unnoticed so far. The second is to try to understand why such a generalization is so pervasive as to characterize persistently widespread forms of dislocation across such diverse languages. It is argued that the generalization in question can be understood in terms of two assumptions: (i) the binding relation involved in SSCO and SWCO is indirect binding (i.e. it involves the so called E-type strategy) and (ii) WCO can be explained on the basis of the idea that traces and pronouns not only have a different syntax, but also a different semantics (i.e. the ‘bimodal’ approach to crossover – Büring 2004, Chierchia 2020). Under this view, the generalization unveiled by Bhatt and Keine reduces to a straightforward application of Principle C. A more speculative conclusion we will offer is that Pinciple C itself may be in turn eliminated in favor of a generalized form of Minimality, a la Rizzi. Comparisons with other proposals (including, of course, Bhatt and Keine’s) will be discussed.