We examine the acquisition of grammatical gender of the definite determiner in Dutch by bilingual children from ethnic minority communities. The results show a complete acquisition of the neuter definite determiner (and demonstrative) by monolinguals (age 9;3-10;5), whereas the bilinguals of the same age perform at chance level. This can partly be explained by the fact that the Dutch nominal gender feature has a default (non-neuter) and a specific value (neuter). Furthermore, these bilinguals are exposed to a Dutch input that is quantitatively and qualitatively different from the standard Dutch input of monolinguals.Because the Dutch input is below a certain threshold, the children's access to the right morphological form regarding grammatical gender of the definite determiner in the storage component is not optimally efficient.
IntroductionGenerally, in the (generative) literature on acquisition, a clear distinction is made between children acquiring two languages from birth (2L1) and children acquiring a second language starting between age 4 and 7 (child L2), although both are sometimes called bilingual 2 children. The most obvious difference between those two groups of child acquirers is the initial state of their emerging grammars.
1Whereas in 2L1 acquisition the two grammars emerge simultaneously, in child L2 acquisition the emergence of the two grammars takes place successively, at least for a large part. There are, however, also important similarities between the two types of acquisition, which, strangely enough, are rarely mentioned in the literature. Most importantly, in both 2L1 and child L2 acquisition the child is supposed to have full access to UG, as opposed to what is sometimes assumed for adult L2 acquisition. In that sense, both 2L1 and child L2 acquisition are more similar to (monolingual) L1 acquisition than to adult L2 acquisition. There may be more similarities since many children acquiring two languages may be situated in a kind of learning situation that has characteristics of both 'pure' 2L1 and of child L2 acquisition, as we will discuss more extensively later with respect to 'our' bilingual subjects. The learning situation is influenced by the sociolinguistic embedding of the second language (and its speakers). For instance, the second language of a 2L1-acquiring child can be the language of the family or the surrounding minority community the child belongs to. Hence, dominance of one of the languages may exist in the input, as may language mixing. Consequently, the sociolinguistic embedding of the second language may have an effect on quality and the quantity of the linguistic input.With respect to child L2 acquisition, the L2 grammar starts to develop at a moment when at least a (large) part of the L1 grammar is already in place. Therefore, although direct access to UG is possible, influence of the L1 grammar on the L2 grammar is to be expected, in the sense that the grammar of one language (L1) may somehow feed the emerging grammar of the other (L2). This influence is genera...