The (In)Stability of Linguistic Systems 1 Running head: THE (IN) STABILITY OF LINGUISTIC SYSTEMS The (In)Stability of Linguistic Systems: Evidence from the Performance of Spanish Monolinguals, Heritage Speakers, and Second Language Learners

The present paper examines (in)stability in the interpretation and production of clauseinitial direct objects in the Spanish grammars of monolinguals, heritage speakers, and second language learners. Analysis of data obtained through two tasks indicates a developmental continuum: monolinguals most closely approximate grammatical targets in which generic DPs are fronted into a Topic position and specific DPs appear in Clitic Left Dislocation constructions; heritage speakers manifest more variable behavior characterized by contraction in the projections that are pertinent to the licensing of fronted objects; second language learners demonstrate a preference for passive and intransitive constructions that confirms a lack of command of the structures of the left periphery, although they display incipient sensitivity to the relationship between bare DPs and Topicalization. The (In)Stability of Linguistic Systems 3 The (In)Stability of Linguistic Systems: Evidence from the Performance of Spanish Monolinguals, Heritage Speakers, and Second Language Learners In recent years, significant attention has been devoted to the interfaces between the core grammar and other peripheral systems, e.g., the perceptual/articulatory and the cognitive/interpretive components. One generalization that emerges from studies couched within this approach is that the narrow syntax, which is controlled by formal features such as Case and Agreement, demonstrates greater stability than the interface with cognitive/interpretive modules, which are regulated by lexicoand discourse-semantic considerations. The present paper pursues this line of inquiry by reference to the interpretation and production of generic and specific DP-fronting in Spanish among monolinguals, heritage speakers, and intermediate adult second language (L2) learners. The presentation is organized as follows. The first section outlines the target of investigation, clause-initial direct objects in Spanish, whose interpretation and production is mediated by the articulation between the clause and the larger discourse. The next two sections present the findings of selected studies on variability in the linguistic performance of heritage speakers of Spanish and adult L2 learners of Spanish, foregrounding efforts devoted to the interface components of the grammatical system. The ensuing section describes the present investigation, its participants, materials, and protocols. The subsequent two sections delineate the results of the study and discuss its broader implications. The (In)Stability of Linguistic Systems 4 The object of investigation: Clause-initial direct objects In Spanish, the ordering of constituents with bivalent predicates is determined by information structure, in particular, the distinction between new information —focus— and old information—topic (Contreras, 1976; Hernánz & Brucart, 1987; Suñer, 1982; Zubizarreta, 1998). The canonical ordering in declaratives is SVO, as shown in (1a) and (2a), but structures encoding non-neutral alternate pragmatic values, dictated by the information structure of the sentence, may be derived by preposing internal arguments to clause-initial position. This is the case in the examples in (1b) and (2b), in which the direct object, thematized in the discourse, is fronted. Note that that in (2b), the fronted element is copied by a resumptive pronominal clitic, a property that will be addressed in the ensuing discussion. (1) a. Nydia pidió un café. Nydia requested-3SG a coffee ‘Nydia requested a cup of coffee.’ b. Un café pidió Nydia a coffee rquested-3SG Nydia ‘A cup of coffee Nydia requested.’ (2) a. José y Yesica hicieron el pastel. José and Yesica made-3PL the cake ‘José and Yesica made the cake.’ The (In)Stability of Linguistic Systems 5 b. El pastel lo hicieron José y Yesica. the cake CL made-3PL José and Yesica ‘The cake José and Yesica made.’ In discussing sentential configurations, researchers such as Rizzi (1997), Poletto (2000), and Benincà (2001), among others, distinguish between the propositional content that is expressed in IP (termed Finiteness) and the ‘articulation of discourse’ (termed Force) expressed by the projection of functional structure, including Topic and Focus, in the area of the left periphery. Of significance for the present discussion are the distinctions among three types of non-Focus object-preposing: Hanging Topics, Clitic Left Dislocation, and Topicalization. Hanging Topics are base-generated in the left periphery, i.e., they are not linked to elements in the IP, and, therefore, are not marked with the case of the resumptive element in the clause: