This article examines the arguments for, and rejects, the proposal by Ackema and Neeleman (2003) that the behavior of the Dutch 2nd person singular pronoun jij in inverted structures should be explained as morphosyntactic allomorphy, conditioned by “initial” prosodic phrasing prior to Spell-Out. First, by neutralizing (under inversion) the distinction between 2sg. and 1sg. present tense verb forms, the proposal makes an incorrect prediction for a well-known class of “strong” verbs. Second, “initial” prosody does not appear to condition the process. Benmamoun and Lorimer's (2006) “overapplication” data for this phenomenon are shown to result from an incorrect interpretation of “d-weakening” verbs.