Complex embeddings in free speech production among late Turkish-Dutch bilinguals

This study investigates potential changes or simplifications in the L1 grammar of late Turkish-Dutch bilinguals in a migrant setting and examines the non-linguistic factors that might have played a role in this process. In our analyses, the bilingual group’s production of embedded sentences in spontaneous speech was compared to that of a monolingual reference group. Based on previous frameworks, complex embeddings were ranked according to their morphological complexity, and it was investigated to what extent a change in preferential patterns (if any) could be explained by linguistic and socio-cultural preferences and demographic characteristics of the bilingual group. Statistical analyses revealed a slight tendency to underuse the most complex types of embeddings by the bilingual group, while their performance did not differ from the monolingual group for any of the other embedding types. Where a difference in performance was observed, the level of education turned out to be the only parameter that influenced the outcome, while language use patterns, age of immigration, length of residence or cultural orientations did not have any impact.