Abstract for Leonie Cornips's talk

Neuter gender and interface vulnerability in child L2/2L1 Dutch: Processing or representation problems?
Leonie Cornips (Meertens Instituut, Royal Netherlands Academy of Sciences)
October 25, 2005 (Tuesday)
6:30 PM - 8:00 PM; Room 7102, The CUNY Graduate Center

In this talk we will consider the gender acquisition of the definite determiner in Dutch. Our subjects are bilingual children aged between 3;0 and 10;5 who are born in the Netherlands and attend a Dutch (pre)school. They belong for the larger part to ethnic minority communities, i.e., these children are from Moroccan, Ghanaian, Turkish and Surinamese families. Strictly speaking, they acquire two languages from birth, in the sense that they got input in both languages from birth onwards. However, it is not entirely clear what the quality and the quantity of this input has been. Sorace (2005) suggests that destabilizing quantitative and qualitative differences in the input to which bilinguals are exposed may affect grammars in different ways: quantitative differences may affect processing abilities, causing delay, whereas qualitative differences may affect representations, causing a qualitatively different grammar, which may eventually lead to language change.

Our results with a sentence completion test (Zuckermann 2001) show that both the bilingual and the monolingual children overgeneralize the non-neuter definite determiner de: they use de instead of the neuter definite determiner het. The L1 youngest and middle age groups, however, go through a significant development of correct use of het, whereas the bilingual children show no significant development of the acquisition of het at all. We will argue that the differences between the monolingual and bilingual children in our experiment are not the result of cross-linguistic influence, but of quantitative and qualitative lack of evidence for the neuter determiner het below a certain threshold in the input of the bilingual children. Although this seems to confirm Sorace's hypothesis, the question can be raised whether this is necessarily a qualitative difference in the sense of a misrepresentation. One could also suggest that because of the 'deficient- input, these children's access to the morphological spell-out of het is not optimally efficient and therefore a competition arises between the default (non-neuter) definite determiner de and the specific (neuter) definite determiner het, which is won by the default de, in most cases.

These two hypotheses make different predictions with respect to the acquisition of other gender-related (agreement) phenomena in Dutch by these same bilingual children.