Relative Scope Of Quantifiers: A Chain-based Approach

Relative scope is a typical manifestation of the displacement feature of natural language. Chains are dependencies between two or more positions of a syntactic structure which are created by the syntactic computation. Relative scope and chain are intuitively related. Our assumption of the Relative Scope Principle (RSP) is related to but different from previous studies on the relation between relative scope and chain. The RSP consists of a semantic/pragmatic condition and a syntactic condition. The semantic/pragmatic condition requires that the indefinite scopal elements must move to the specifier position of a functional category XP to check off their ?-features. The syntactic condition requires that relative scope appears if and only if the heads of the chains of the indefinite scopal elements co-c-command each other. The empirical evidence to support our hypothesis of the RSP focuses on the linguistic data in literature of A-movement reconstruction. The concerned literature argues against the existence of A-movement reconstruction by studying intensively the interaction between quantifiers and negation and the difference between the raising construction and the control construction. We re-examine these data with reference to the RSP. In the analysis of negation, our discovery is that the constituent negator modifying a quantifier never triggers the movement of the concerned quantifier up to [Spec, XP] and it never induces relative scope reading. In the analysis of raising and control constructions, we find that when a quantifier and its co-indexing elements are in separate chains, the quantifier has the definite reading and needs not move up to [Spec, XP]. We also find that when a quantifier and its co-indexing element are in a single chain, the quantifier has the indefinite reading and must move up to [Spec, XP]. It is in this way that the phenomenon where the concerned control construction always induces the relative scope while the concerned raising construction sometimes does. The conclusion drawn from our analysis within the RSP confirms with Chomsky's idea of the non-existence of A-movement reconstruction.