CRL Talks

This presentation reports on a multi-study PhD project investigating syntactic processing in Norwegian-English bilinguals, with particular focus on code-switching, word order, and grammatical agreement. The project examines how bilingual grammars handle structural conflicts between Norwegian's Verb Second (V2) requirement and English's Subject-Verb-Object (SVO) word order, moving from group-level judgments to individual differences and morphosyntactic processing.
Study 1 employed an online acceptability judgment task (N=43; 29 Norwegian-dominant, 14 English-dominant bilinguals) to investigate preferences for V2 versus V3 word order in clause-initial adverbial contexts across unilingual and code-switched sentences. Preliminary findings reveal that word order preferences consistently track the matrix language: both groups prefer V2 in Norwegian-headed clauses and V3 in English-headed clauses. Crucially, English-dominant bilinguals show significantly higher acceptance of V3 structures in Norwegian matrix clauses (p = .037), suggesting that strong English exposure increases tolerance for English-influenced word order even when Norwegian provides the syntactic frame. 
Study 2 (in development) will use self-paced reading methodology with Norwegian-dominant English bilinguals to investigate individual differences in online processing costs for V2 versus V3 structures. Using refined stimuli based on the acceptability study and corpus data, this experiment shifts focus from group comparisons to individual variation, examining how factors such as language experience, dominance patterns, and cognitive control relate to processing costs during code-switching. We hypothesize that individual differences in cross-linguistic influence and working memory capacity will modulate acceptance and processing of non-canonical word orders.
Study 3 (planned) will examine gender agreement processing within Norwegian L1 using filler sentences from the self-paced reading experiment. Norwegian exhibits grammatical gender (masculine, neuter, and in some dialects feminine), while English lacks gender marking. We will investigate processing costs when gender-marked indefinite articles mismatch with noun gender, using corpus-verified stimuli with unambiguous masculine and neuter nouns alongside mismatch conditions.