ABSTRACT
Chapters 3 and 4 both argued that certain phenomena that had previously been analyzed in terms of covert movement should be reanalyzed in terms of overt movement. Given a condition like Procrastinate, overt movement must be forced by some requirement, a requirement that Chomsky has stated in terms of “strong features.” This chapter compares several possibilities that Chomsky has suggested for the condition demanding that strong features must be checked in overt syntax. Based on the analysis of Pseudogapping of Chapter 5, I show that ellipsis, which I analyze as deletion, can “repair” a structure damaged by failure of normally obligatory overt movement (V raising, in this case) to apply. I then give a parallel analysis of Sluicing, with corresponding implications for Infl raising to C. I show how these analyses fit neatly into the Chomsky (1993) PF crash theory of strong features, although only if the strong features in the constructions at issue reside in the item that is to move, rather than in the “attracting” head (the latter being Chomsky’s position on the issue). I then proceed to show how the repair phenomenon can actually be reinterpreted to bring it into line with Chomsky’s position on the locus of feature strength. However, I conclude with one phenomenon that seems to argue powerfully against the claim that strong features are invariably in the attracting head: multiple wh-fronting in Slavic. Why should the attracting head need more than one wh-phrase? Interestingly, Bosˇkovic´ (1999) now directly addresses this question, positing an “Attract-all” feature, thus preserving the possibility that Chomsky’s position is correct.
