Miriam Butt (University of Stuttgart) & Tracy Holloway King (Stanford University)
The strong claim put forward by Horvath (1995) that structural focus is licensed by mechansims which parallel structural Case assignment receives support from SOV languages like Urdu and Turkish. However, the case marking, position, and interpretation of objects do no yield the expected straightforward parallelism to focus licensing. In particular, a conflict arises with regard to the interpretation of the immediately preverbal position.: while preverbal focus forces a specific interpretation, unmarked preverbal objects are interpreted as nonspecific. Following Horvath, focus is associated with a functional head (Infl) and a lexical item (V) must raise to license focus; focus in these languages is assigned under government and adjacency, which is satisfied by right-adjunction at AspP, creating a preverbal position. In Turkish and Urdu, nominative objects can only be interpreted as nonspecific in immediately preverbal position; this adjacency requirement follows from Weak Case being assigned to the complement of V and forcing a nonspecific reading. Since focus requires V-raising, our account correctly predicts that WEak Case objects cannot occur with preverbal focus phrases; nominative objects which co-occur with preverbal foci always receive Strong Case and the corresponding semantic interpretation. Thus, we show that a unifying analysis for the crosslinguistic similarity of focus and Case follows from a combination of Horvath's general approach and the differentiation between Weak and Strong Case (de Hoop 1992).