Neither agreement nor pronouns: Clitic doubling and Weak Crossover in Rioplatense Spanish

Abstract

Baker & Kramer (2018) suggests that the distribution of clitic doubling in Spanish follows from Weak Crossover. Thus, the fact that in accusative clitic doubling bare wh-phrases cannot be doubled (e.g., *¿A quién lo viste? ‘Who did you see?’) is explained by the same reason that explains a standard WCO violation (e.g., *?¿A quién vio su madre? ‘Who did her mother see?’). If this in on the right track, then, accusative doubling clitics must be considered plain pronouns. The fact that the distribution of dative clitic doubling is wider than accusative clitic doubling, allowing, for instance, the doubling of bare wh-phrases (e.g., ¿A quién le diste un libro? ‘Who did you give a book?’), is, consequently, taken by Baker & Kramer as evidence that dative doubling clitics cannot be pronouns, but mere agreement markers. In this reply, I show, mainly based on data from Rioplatense Spanish, that both conclusions are incorrect. Baker & Kramer’s suggestion regarding accusative doubling both overgenerates and undergenerates. What regulates the distribution of doubling clitics in Rioplatense Spanish and beyond is the inflectional makeup of objects. Concretely, direct objects are doubled whenever they bear a [person] feature (Di Tullio et al 2019), whereas indirect objects are doubled by the mere presence of a [D]-feature (Pujalte & Saab 2018). Yet, despite this difference, both are probes for A-movement and predicate-makers at LF, i.e., neither plain pronouns nor mere agreement markers.