Recent research posits the role of processing mechanisms on cross-linguistic influence (CLI) (Vasilyeva et al., 2010; Nicoladis, 2012; Hervé et al., 2015). However, hardly any study considers CLI in simultaneous bilingual children during real-time language processing (online). The aim of this paper is to (i) examine how online vs. offline processing, and (ii) individual measures of both fluency and language exposure affect CLI at the determiner level in French-English children. French requires the projection of a determiner in generic context when English allows bare mass nouns and bare plural nouns.
Forty-five 8-to-10-year-old French-English children took part in a self-paced reading (SPR) task, an Acceptability Judgement task (AJT) and a Cloze-test in their two languages. French and English monolinguals (n = 24) were also included. A parental questionnaire measured language exposure and fluency (Cattani et al., 2014).
Unlike previous studies (Hervé & Serratrice, 2018; Kupisch & Pierantozzi, 2010; Serratrice et al., 2009), our results provide evidence of bi-directional CLI. In English, CLI occurs only in the cloze- test. Exposure to English, age and noun type distinction affect the bilinguals’ performance. In French, CLI is evidence only in the offline AJT but not in the cloze-test nor in the SPR task. In the French AJT, the children exposed the least to English performed significantly better. Rating accuracy also increased with age. These results are discussed in terms of implicit and explicit knowledge.