Francisco Ordóñez

Left dislocated subjects and the pro-drop parameter: A case study of Spanish☆

Abstract The typical assumption in a pro-drop language like Spanish is that covert as well as overt subjects occupy a preverbal position at Spell Out in which their case and agreement properties are satisfied. This paper presents evidence against such a claim. On the one hand, we show that pre-verbal overt subjects pattern with left dislocated DOs and IOs in a wide range of syntactic contexts: ellipsis, extraction of quantifiers and interpretation of preverbal quantifiers. In these same contexts, sentences with a silent subject differ from sentences with overt ones. We conclude that overt pre-verbal subjects are necessarily left dislocated. In order to account for the left dislocated nature of overt subjects, we propose to eliminate AgrS as a functional projection. Instead we take the idea that subject agreement should be considered a clitic (Taraldsen, 1992), and the relation between the agreement and subject to be one of clitic doubling. Evidence in favor of this claim comes from striking parallelisms between standard clitic doubling constructions and agreement-subject constructions. Specifically, both cases pattern similarly in relation to the determination of binding in certain cases of mismatches in person between the doubling DP and the clitic. Since we take agreement to be a clitic that absorbs theta role and case, movement of the doubling DP subject to a preverbal position cannot be driven by agreement or case reasons. Instead, movement of the subject to a pre-verbal position is driven by discourse considerations as is typical in left dislocations.

Post-Verbal Asymmetries in Spanish

Spanish allows its post-verbal subjects to appear in the VSO and VOS configurations. It has been generally assumed that the second order is generated by the adjunction of the subject to the right (Rizzi 1982, Torrego 1984, and Sun_er 1994). This paper explores an alternative to this traditional view in which this order is generated by the scrambling of the objects to the left. Empirical support in favor of this hypothesis comes from certain syntactic asymmetries between VSO and VOS. Some of these asymmetries reflect the fact that the object c-commands the subject in the VOS order but not in the VSO order. In other cases, the asymmetries show that certain types of objects cannot move to the left to produce the VOS order. This is the result predicted by the constrained nature of scrambling. Specifically, there is a parallel between these alternations in Spanish and the same ones described in scrambling languages (e.g., German or Korean) with SOV and OSV alternations. Finally, this hypothesis supports the line of research put forward by Kayne (1994) for word order in UG, according to which right adjunction is not possible.