Reconstruction Effects of Binding Conditions
摘要
This chapter addresses an asymmetry between coreference and bound variable anaphora with respect to Condition C reconstruction. Following Higginbotham’s (Linguistic Inquiry 14:395–420, 1983) claim that strong crossover (SCO) cases cannot be treated on a par with typical cases of Condition C violations, it is argued that the reconstruction effects of typical Condition C violations are captured by way of the timing of lexical overlay onto pro that undergoes movement to establish an anaphoric relation. On the other hand, the various SCO cases that Higginbotham (Linguistic Inquiry 14:395–420, 1983) accounts for in terms of his chain condition are captured under our movement theory by modifying the notion of “minimal” relevant to minimal Search. Further, I argue for the position taken by Safir (The syntax of (in)dependence. MIT Press, Cambridge, MA, 2004) and Barker (Linguistic Inquiry 43: 614–633, 2012) according to which in order to function as variables bound by wh-phrases and QPs, pronouns need to obey the scope requirement rather than the c-command requirement. This chapter also deals with interesting patterns of binding reconstruction in the parasitic gap (PG) construction: One pattern shows that such reconstruction does not take place into PGs whereas the other pattern shows that such reconstruction is mandatory. It is argued that such apparently contradictory patterns are explained as a result of the interaction of our movement theory with the proposal that the PG construction involves not only a null operator, as proposed by Chomsky (Barriers. MIT Press, Cambridge, MA, 1986), but also what I call a copy operator, i.e., a copy of a real operator.