Accelerated Acquisition in Spanish-English Bilinguals: The Structural Transfer Hypothesis

The bilingual child does not fuse the two languages to which she is exposed; rather, in each language, most of the milestones she reaches, and even most of the errors she makes, align neatly with those of a monolingual acquiring that language. Hence it is generally agreed that the bilingual child acquires two separate grammars (e.g., Meisel, 1989, Genesee et al., 1995). Yet this separation does not preclude some influence between the two grammars as they develop, such that bilinguals sometimes acquire certain constructions earlier or later than monolinguals do (e.g., Hulk & Müller, 2000), and on rarer occasions bilinguals produce types of utterances not ever attested in monolinguals’ speech (e.g., Döpke, 1998; Yip & Mathews, 2000). Understanding cross-linguistic influence in the domain of syntax requires the identification of the conditions under which one developing grammar influences the development of the other grammar. One prominent proposal, that of Hulk and Müller (2000), stakes out two conditions required for negative cross-linguistic influence, which consists of a slower or different pattern of acquisition in comparison with monolingual norms. That proposal states that (i) cross-linguistic interference occurs at the syntax/pragmatics interface because of the challenge this presents even in monolingual acquisition, and that (ii) for interference to occur, surface structures must be sufficiently similar that they allow for an initial confusion of underlying forms. Other approaches have stressed that Hulk and Müller’s two-constraint formulation on interference is deficient when additional language pairs or later stages of syntactic development are examined (e.g., Serratrice et al., 2004; Notley et al., 2007); in such instances, researchers suggest that other factors such as input frequency, transparency of syntaxpragmatics mapping, complexity of target structures, and Chomskyan economy principles might be usefully considered. Less commonly attested but theoretically consequential is cross-linguistic influence that accelerates bilinguals’ acquisition of certain structures when compared with monolingual norms (e.g., Kupisch, 2005). Given the diversity of opinions on this subject, it is difficult to make predictions about whether, when, and what kind of influence will occur for any language pair. Against Hulk and Müller’s Interference Hypothesis predicting only negative cross-linguistic influence I present the Structural Transfer Hypothesis (STH), which supplies conditions for three types of cross-linguistic influence in bilingual first-language acquisition (BFLA): ‘acceleration’, in which bilinguals acquire a given construction earlier in development than monolinguals; ‘delay’, in which bilinguals acquire a given construction later in development than monolinguals; and ‘interference’, in which bilinguals pass through stages of linguistic development that monolinguals do not. The object of empirical investigation is the acquisition of wh-questions in Spanish-English (S-E) bilinguals, selected because these questions are known to pose difficulties in the monolingual acquisition of English (§2.1) and also engage the notoriously challenging C-domain, suggesting the possibility of interference (§2.2). Corpus analyses reveal that the acquisition of wh-questions in S-E bilinguals is in fact accelerated (§3), a phenomenon explained by the STH (§4) when taken in light of a particular interpretation of Spanish syntactic development (§4.1). Finally, the STH is shown to capture a variety of instances of cross-linguistic influence found in the BFLA literature (§4.2).