Abstract
The availability of preposition omission in ellipsis is assumed to follow from the possibility of preposition stranding in corresponding non-elliptical clauses (Merchant 2001). To test this assumption I sampled 409 ellipsis remnants from three corpora of American English, and developed generalized mixed-effects models of preposition omission in ellipsis. The results show that preposition omission is, among others, sensitive to gradient constraints independently found to influence preposition placement in interrogative clauses, without being correlated with the possibility of preposition stranding. This finding challenges evidence for syntactic parallels between elliptical constructions and their nonelliptical counterparts, and hence, for deletion-based approaches to ellipsis.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
