Ackema and Neeleman (2003) discuss three phenomena that arise in the context of agreement and pronominals: agreement asymmetries, cliticization, and null subjects. They develop a unified analysis for these phenomena, claiming that they all involve a process of weakening within prosodic domains. While we agree with their important insight that the PF interface is responsible for some of these phenomena, we will argue against their weakening analysis. We provide arguments that agreement asymmetries cannot be uniformly analyzed as involving the same processes as phonological cliticization or null subjects. We instead propose that the observed asymmetries arise because of the alternative forms of spelling out features at the PF interface.

Aoun, Benmamoun, and Sportiche (ABS, 1994) propose an analysis of first conjunct agreement in VS sentences in Lebanese Arabic and Moroccan Arabic. On the basis of the distribution of number-sensitive items, they argue that this type of agreement is due to clausal coordination. Munn (1999) argues against ABS's account and proposes that first conjunct agreement in the Arabic dialects arises because coordination of NP subjects is semantically plural but syntactically singular. In this reply we show that Munn's alternative analysis is empirically inadequate.

The Arabic quantifier kull displays a Q___NP and NP___Q alternation. Shlonsky (1991) argues that in both patterns Q heads a QP projection with the NP as a complement that may undergo movement to [Spec, QP] or beyond to yield the NP___Q pattern and Q-float structures. On the contrary, I argue on the basis of evidence from reconstruction, Case, and agreement that the two patterns are radically different. In the Q___NP pattern Q is indeed the head of a QP projection that contains the NP. In the NP___Q pattern, however, Q heads a QP adjunct that modifies the NP and in some cases the VP.

We investigate the interaction of clitic left-dislocation (CLLD), wh -interrogatives, and topicalization in Lebanese Arabic. A wh -phrase or a topicalized phrase can be fronted across a CLLDed element derived by movement but not across a base-generated one. A CLLDed element cannot be fronted across another CLLDed element, a wh -phrase, or a topicalized phrase. These interception effects are accounted for only if Minimality is construed as a constraint on derivations rather than representations and if fronting of the CLLDed elements is seen to apply in the PF component. It is thus suggested that the mapping between overt Syntax and the Articulatory-Perceptual level is not trivial.