Under the influential Roothian proposal for focus association, focused phrases remain in-situ at LF (Rooth, 1985, 1992). However, a recent line of work has resurrected the idea that focus association involves covert movement: specifically, the associate of English sentential only must covertly move to only (Drubig, 1994; Krifka, 1996, 2006; Tancredi, 1997, 2004; Wagner, 2006). Hadas Kotek and I have developed new arguments for this idea that association with English sentential only involves covert focus movement, with the possibility of pied-piping. We show that covert pied-piping in focus association explains previously unnoticed patterns of island-sensitivity in Tanglewood constructions (Kratzer, 1991) as well as patterns of focus intervention effects in focus constructions. We also show that covert focus movement can feed reflexive binding and parasitic gap licensing.

Return to all projects