No case tampering once transfer domain is formed!

Abstract This research article offers empirical evidence from Standard Arabic (SA) that an existing structural case assigned on an element by one head can be overridden by a new structural case assigned by a different head as long as the element (or one of its copies) has not become part of any previous transfer domain defined by the Phase Impenetrability Condition (PIC) (see Chomsky, Noam. 2000. Minimalist inquiries: The framework. In Roger Martin, David Michaels & Juan Uriagereka (eds.), Step by step: Essays on minimalist syntax in honor of Howard Lasnik, 89–155. Cambridge, MA: MIT Press). Our main evidence comes from the patterns of case assignment of the overt complementizer ʔinna in SA. ʔinna can only assign case to elements that otherwise bear default case, including a topical object or a topical subject as well as elements that are assigned case by T0 such as a contrastively focused subject. On the other hand, ʔinna never assigns case to a contrastively focused object (that is located in CP) which is argued to be base generated in its thematic position within the transfer domain of v*P. These facts are taken together as evidence that a structural case assigned to elements within a phase is temporary (as it can be overridden) until the transfer takes place. We attribute this to the workings of a transfer principle labelled as The Case-Chain Uniformity Principle (CCUP) that demands that a non-trivial chain (i.e., a discontinuous entity) be only assigned one case value in the interface.