Rightward Movement and Ellipsis in the Syntax of Spanish

It has been argued that a variety of ellipsis constructions in Spanish show properties of rightward movement, e.g., Wh-Stripping (ellipsis with a wh-remnant other than why followed by a non-wh-remnant, [1]), (1a), Multiple Sluicing (Sluicing with multiple remnants, [2] following [3]), (1b), and Gapping (an ellipsis construction where a verb goes missing in the second conjunct of a coordinate structure, [4], a.o.), (1c). Goal: to capture the fact that i. rightward focus movement does not apply in non-ellipsis contexts; ii. rightward focus movement arguably does not apply in ellipsis structures involving a single remnant, e.g., Sluicing; iii. rightward focus movement is restricted to a subset of ellipsis structures involving multiple remnants, e.g., it applies in the structures in (1), in contrast to Why-Stripping, which typically involves the wh-element why (por qué in Spanish) followed by a non-wh-phrase, (2). Analysis: Following [5] and references therein, evidence is provided for the view that in the structures in (1) there is full-fledge syntax at the ellipsis site, as shown by various connectivity effects, and (ii.) the remnant at the right edge shows properties of rightward movement (see [6], a.o), as shown by the fact that it is clause-bound and that it cannot drop the preposition, (3a) and (3b), respectively (data exemplified for Gapping; see [7]), in contrast to Sluicing; see (4) for illustration of the P-stranding facts in Sluicing ([8]). Under the assumption that derivational antisymmetry ([9]) bans multiple Specifiers due to their not being linearizable, it is argued that ellipsis provides a way to solve this linearization conflict. Specifically, ellipsis creates an escape hatch for multiple remnants competing for the same position when attempting to survive ellipsis, e.g., in the structures in (1) as opposed to Why-Stripping, where why is generated higher than the landing site of the focused XP and, therefore, there is no competition among the remnants, [10]). It is claimed that in the former constructions one of the Specifiers may take the place of the elided chunk of structure at PF, resulting in what looks like rightward movement. This means that specific aspects of the LCA in its more recent instantiations ([9]), in particular, the ban on multiple Specifiers, can be violated under well-defined circumstances (see also [11]). Accordingly, an analysis is developed in OT-terms, as this framework is particularly well-suited to handle the interaction among multiple factors. Inasmuch as no competition is involved in Why-Stripping, this structure patterns …