A-Chains and the Copy Theory
Rizzi’s chapter starts by discussing the functional motivations of movement (i.e., movement as a last resort) and then connects this level of analysis to the study of the form of chains—in particular, A ¯-chains—and the formal principles that constrain possible chain configurations. Rizzi proposes that A ¯-chains connect two semantic interface positions: an s selection position and a criterial position, which expresses some scope discourse property. He further tries to determine whether and under what conditions other positions are allowed to occur in well-formed chains, be sides these semantic interface positions. He provides evidence for a principle of Criterial Freezing: a phrase meeting a criterion is frozen in place, and its chain cannot extend further. Rizzi further investigates subject chains, especially the EPP property (the obligatory appearance of subjects in clauses) and the Empty Category Principle property (the impossibility of subject extraction). He reduces the EPP to a particular criterion, the Subject Criterion, and proposes that the immobility of subjects can be analyzed as a consequence of Criterial Freezing.
In her chapter, Yang argues that two apparently related constructions, relative clauses and questions, do not necessarily involve the same movement operation. She argues that relative clauses in both Hindi and Punjabi involve successive-cyclic A ¯-movement, while questions in both languages involve scrambling. Her argument is based on the observation that in Punjabi, an intermediate copy of a relative phrase can be overtly realized in Spec,CP. Furthermore, more than one copy in a relative movement chain can be pronounced. She argues that the multiple realization of relative phrases in a movement chain can only be accounted for under a theory of multiple Spell-Out (such as that proposed in Bobaljik 2002). She further shows that the relevant facts concerning multiple realization fall out from the CP-Phase Constraint, according to which only one copy of a movement chain can be overtly realized in a given CP phase.
Reintges, LeSourd, and Chung begin their chapter by contending that the copy theory of movement allows for the possibility that the lowest copy is pronounced. They argue that certain types of wh-in-situ patterns instantiate ‘‘hidden’’ movement configurations that are created by wh movement in overt syntax with subsequent pronunciation of only the lowest copy. Their evidence comes from the phenomenon of wh agreement: internally headed relative clauses in Passamaquoddy, an Eastern Algonquian language, and wh-in-situ interrogatives in Coptic Egyptian, an Afroasiatic language spoken in the fourth through the fourteenth centuries ad. Under the assumption that this special wh-agreement morphology records the history of wh-movement before the derivation is shipped to PF, its occurrence in certain wh-in-situ constructions reveals that in these cases as well, movement has taken place in overt syntax. The only difference from overt wh-movement is that the lower rather than the higher copy is pronounced. The authors also show that despite superficial similarities, wh-in-situ constructions in Palauan, an Austronesian language, contrast sharply with Passamaquoddy relatives and Coptic wh-in-situ interrogatives. While wh-agreement is found in Palauan questions formed with overt wh-movement, this agreement is not triggered by wh-in-situ constructions, which suggests that no narrow syntactic movement is involved in such constructions. The authors propose that in those cases, the scope of the interrogative phrase results from unselective binding.
