Leísta Spanish and the Syntax of Clitic Doubling

This dissertation introduces clitic doubling data from Leísta Spanish (a dialect spoken in the North of Spain). In this dialect, the dative form clitic is used as a direct object clitic when the referent (or associated overt or covert NP) is animate (sometimes also restricted to masculine). Like other doubling dialects, Leísta Spanish shows a (doubling) asymmetry between direct objects and indirect objects. Direct object doubling manifests animacy and specificity restrictions that do not hold of indirect object doubling. In Chapter 2 I show that this data presents a problem for past analyses of clitic doubling which tie these interpretive restrictions to a particular clitic form. As a first step to solving the interpretive puzzle, I argue that dative and accusative clitics should receive different analyses. Dative clitics are agreement markers (a la Sportiche 1993); accusative clitics are determiners (see Torrego 1988; Uriagereka 1988). The questions raised by Chapter 2 are: (1) How should dative-form accusative clitics in Leísta Spanish be analyzed? and (2) What is the internal structure of the direct object in a clitic doubling construction? These questions are answered in Chapter 4, where I argue that clitic doubling in Leísta Spanish should be analyzed as an instance of possessor raising. In both clitic doubling and possessor raising, a DP-internal constituent raises to be in a spec-head relationship with a dative clitic. Chapter 3 lays the groundwork for this analysis. Here, the internal structure of possessor DPs is established, and it is demonstrated that movement of a possessor DP out of a direct object DP is possible. Chapter 5 addresses the question of the relationship between the prepositional accusative marker a and clitic doubling, known as Kayne’s Generalization. The generalization is that clitic doubling cannot occur without amarking. I explore two possible accounts for this relationship. The first possibility is that the relationship is syntactic. The a-marked element is a dative marked DP internal possessor. Given the hypothesis that clitic doubling involves the raising of a DP-internal possessor, if that possessor is necessarily a-marked, it follows that clitic doubling implicates a-marking. The second hypothesis is that the relationship arises due to the independent semantic properties of the two constructions, and that there is no direct dependence of one on the other. On this view, the semantic properties associated with a-marking are a subset of those associated with clitic doubling. The apparent dependence, then, is due to this subset relationship. Neither of these hypotheses is without problems, and these are discussed in the chapter