Knowledge of Noun-Drop across Various Lexical and Functional Categories in Heritage Spanish Bilinguals

Recently, language acquisition research from the generative paradigm has turned its attention to investigating heritage language bilingualism in the hopes that studying the adult competence outcomes of these childhood bilinguals can inform key issues pertaining to bilingual acquisition more generally as well as issues relating to L1 attrition and adult L2 acquisition theorizing (see Montrul 2008a; Rothman 2007, 2009). Heritage speaker bilingualism occurs under specific environmental circumstances. Crucially, a language qualifies as a heritage language if it is a minority language in its context as a heritage language and is the home language of the heritage speaker acquirer. Under such a scenario, the heritage language, like the case of monolingual acquisition, is acquired naturalistically at a very young age on the basis of an interaction between primary linguistic data and innate linguistic principles. More often than not, it is the sole first language of the heritage speaker despite the fact that in adulthood heritage speakers are likely to show deficits in knowledge as compared to monolingual controls. Understanding the variables that conspire to explain the competence differences between these childhood bilingual and monolingual adults, in addition to posing interesting new research questions specific to heritage language bilingualism, innovatively informs the age of exposure/acquisition debate and helps researchers choose between opposing approaches (see Montrul 2008a). language in the United States has demonstrated that heritage speaker competence in adulthood is selectively different from that of monolinguals, effectively dispelling the notion that language acquisition in childhood is virtually synonymous with complete convergence and questioning the extent to which the age of exposure debate is properly informed or contextualized. Another general research program to which heritage speaker acquisition studies contributes significantly is the so-called Interface Vulnerability Hypothesis (IVH), which claims that narrow syntactic features are acquired unproblematically whilst the acquisition of properties that require integration of information across cognitive-linguistic modules, otherwise known as interfaceconditioned properties (in the sense of Jackendoff 2002 and Reinhart 2006), remain susceptible to incomplete acquisition, attrition or delays in full convergence depending on the type of learning scenario (e.g. Belleti, Bennati and Sorace 2007, Sorace 2005, White to appear). Although recent accounts hold such vulnerability to external interfaces (the syntax-pragmatics interface) as opposed to internal ones (the syntax-semantics interface) based on the fact that the latter actually results in ungrammaticality as opposed to gradient levels of infelicitousness like operations at the syntaxdiscourse interface and the former requires a higher processing load (Tsimpli and Sorace 2006; Sorace and Serratrice 2009), the IVH claims that linguistic properties dependent on the discourse interface would be the primary locus for incomplete acquisition and L1 attrition in heritage bilingual grammars. Alternatively, grammatical properties that rely strictly on the computational system should be acquired in a straightforward manner for all language acquirers, including heritage language bilinguals, provided that the input they receive provides the necessary syntactic triggers. The general position of IVH is supported by the fact that in child language, monolingualism and bilingualism, interfaces, particularly the syntax-pragmatics interface, cause delays in acquisition (Platzack 2000; Grinstead 2004; Paradis and Navarro 2003; Serratrice, Sorace and Paoli 2004) or are loci for cross-linguistic influence (be that delays or bootstrapping, see Müller and Hulk 2001, Cornips and Hulk 2006), with