We argue for the existence of covert focus movement in English focus association. Primary evidence comes from Tanglewood configurations of the form in Kratzer (1991). We show that Tanglewood configurations are island-sensitive, which is unpredicted by Kratzer’s focus-index proposal. We propose that Tanglewood configurations always involve covert movement of the overt focus—possibly with pied-piping (Drubig, 1994; Krifka, 2006; Wagner, 2006; Erlewine and Kotek, 2014)---to bind a bound variable in the ellipsis site. We show that covert focus movement is different from QR and can be long-distance.
See “Focus association by movement: Evidence from Tanglewood“ in LI.