Unifying Agreement and Agreement-less Languages ∗

In languages that lack morphological agreement, focus plays a role which is computationally equivalent to agreement. The EPP on T picks out the focused XP and raises it to the Spec of TP, very much like the agreeing phrase in languages with agreement. Why should there be this ‘parametric variation’ between agreement and focus? Following Chomsky (2005), I assume that agreement as well as focus originates on the same head — the phase head C. Either agreement or focus percolates down to T. If focus percolates down, we have a focus-prominent language, while percolation of agreement leads to an agreementprominent language. I will show that focus movement at the level of T resembles wh-movement to C in Indo-European. I suggest that focus movement in focus-prominent languages and wh-movement in agreement languages are essentially the same phenomenon (cf. Miyagawa 2001, in press), a point which accounts for an observation by Fukui (1988) and Kuroda (1988) that only agreement languages have wh-movement.