Topic: HUMAN SEMANTICS
Pseudo Weak Crossover in French relative clauses &
global economy
byBenjamin Spector
(1) Le/( ?)Aucun type que(1) son(1) p?re a frapp? t(1) a pleur?.
The/ No guy that(1) his(1) father has beaten t(1) has cried.
(2) *Le/*Aucun type dont(1) le p?re t(1) l(1)' a frapp? a pleur?.
The/No guy of-whom(1) the father t(1) him(1) has beaten has cried.
Both (1) and (2) are WCO violations, since in both cases the trace does not c-command the pronoun; yet (1) is fine but (2) isn't. Note that (2) would be acceptable with another co-indexing.
(...)The hypothesis I will defend in this paper is similar in spirit to a proposal put forward in Ruys (1994): among the structures that could realize the logical form in question, the most economic one, in a sense to be defined, is preferred. More precisely, grammar generates a set of structures corresponding to a given logical form; a metric is defined over these structures, and the most economic structure is preferred to all others. A crucial point here is that this comparison process will apply only to structures that are otherwise wellformed, i.e. which, for instance, do not violate any known locality constraint. An immediate prediction that is made is that when one of the two potential candidates cannot be generated by the grammar, then there will be only one candidate, which will therefore be selected as the most economic one.
Download link
Related post
Posted by Tony Marmo
at 01:01 BST
