In this article, we claim that syntactic objects undergoing ellipsis can be targeted by both narrow syntactic and PF operations. We base this conclusion on experimental evidence from the interaction between single conjunct agreement and verb-echo answers in South Slavic, which we show to be derived via verb-stranding VP-ellipsis. Adopting the view that Vocabulary Insertion replaces Q-variables on lexical heads (Halle 1991) and that ellipsis is a syntactic operation that deletes Q-variables (Saab 2022), we demonstrate that constituents properly included in the ellipsis site can undergo Internal Merge in the narrow syntax and can participate in PF processes from the derived position. The interaction between ellipsis, Internal Merge, and Agree-Copy that accounts for these patterns of data follows naturally within the Distributed Ellipsis approach.

We discuss cases of salvation and non-salvation by deletion in the domain of lexical gaps, and distinguish two types of defectiveness: (a) defectiveness that can be saved by PF deletion, which we take to signal the lack of an eligible allomorph for certain environments within a language, and (b) defectiveness that cannot be saved by PF deletion, which we take to signal the lack of a proper alloseme for a given environment. With ellipsis modeled as an instruction for nonpronunciation on the PF branch of the grammar, only gaps on the Exponent List can be saved by it.

Harris and Halle (2005) present a framework (Generalized Reduplication) that unites the treatment of phonological reduplication and metathesis with similar phenomena in morphology, thereby accounting for the apparently spurious placement of the imperative plural - n in mesoclitic Spanish forms such as hága-lo-n ‘Do it!’, in which clitic lo is sandwiched between the verbal stem and the plural suffix. Kayne (2010) has challenged their analysis, arguing that such cases should be treated purely within the syntax. We reassess some of Kayne’s arguments, agreeing with his conclusion that the most important desideratum of any general analysis of such phenomena is restrictiveness. However, we contend that greater restrictiveness can be achieved through morphotactic constraints and repairs in the Generalized Reduplication formalism, triggered by a Noninitiality constraint on the positioning of the plural affix, and we develop constraints on these operations that situate interspeaker variation within the postsyntactic component.

This article discusses morphological markedness as a trigger and target of postsyntactic feature-deletion operations ( impoverishment ; Bobaljik 2003, Bonet 1991, Halle 1997, Halle and Marantz 1993, Harley 2008, Noyer 1992, 1998) and, taking number as a case study, argues that dual is more marked than plural, in accordance with traditional and more recent approaches to inflectional morphology. In a system that employs abstract binary features, dual may be represented by a combination of the features [−singular, −augmented] (Conklin 1962, Noyer 1992), and the feature [−augmented] is marked in the context of [−singular]. This article draws a formal distinction between markedness-targeted impoverishment and markedness-triggered impoverishment, arguing that the latter is an important diagnostic for morphological markedness. Exemplification comes from syncretisms either directed at or conditioned by the dual in Sámi, Sorbian, Slovenian, Warlpiri, and Zuni, the last of which has been argued by Cowper (2005) to show that dual is less marked than plural.