Recipient (sub)extraction: a case for cyclic linearization

There is variation between dialects of English in whether or not a recipient argument of the double object construction can be extracted via Ā-movement. I argue that the Ā-extractability of the recipient is contingent upon the phasehood of the Applicative Phrase (ApplP), which introduces it: when ApplP is a phase, it cannot undergo Ā-movement in any dialect because it causes a linearization problem within Fox and Pesetsky’s ( 2005 ) architecture for linearizing syntactic structure. In the dialects where the Ā-extraction is forbidden, phasal ApplP is distinguishable from non-phasal ApplP on the surface. However, in the Lancashire dialect of British English, which allows it, the superficial difference between the two kinds of ApplPs is neutralized, due to an independent factor in the dialect. It gives us the false impression that Ā-extraction of the recipient is permitted even when ApplP is a phase. In the Lancashire dialect, the otherwise unrecognizable phasal status of ApplP can be detected by movement of a theme argument to the edge of ApplP. Since only phasal ApplP can induce this movement, Ā-extraction of the recipient becomes impossible when there is a syntactic indication of movement of this sort.