Comprehension of grammatical gender, case and wh-questions in Greek heritage children

Abstract

Over the past twenty years, one of the most debated questions in bilingual acquisition is how heritage language speakers acquire their heritage language. In this thesis, we address how Greek heritage children acquire their heritage language. The heritage language is the first language (L1) children begin acquiring from birth that corresponds to a minority language. Gradually in the course of language development, the heritage language is taken over from the majority language of the environment where bilingual children grow up. Eventually, the majority language becomes the heritage children’s second dominant language (L2). Under this language contact situation, the grammar of the heritage language in children is characterised by linguistic variation and change.

The aim of this thesis is to explore the morphosyntactic features that are vulnerable (susceptible to change) in Greek heritage language acquisition as they are affected by contact with the majority language English. We argue, first, that ambiguous and opaque morphosyntactic features, such as case, qualify for a vulnerable structure in the acquisition of a heritage language. Second, we look at how factors like proficiency and exposure to the heritage language as well as the age of onset of the L2 contribute to the linguistic change of the heritage language.

The present thesis addresses these questions in three experiments examining the acquisition of the nominal inflectional morphology in a variety of constructions in Greek heritage language children with English as their dominant L2. Study 1 (Chapter 2) investigates gender identification of gender inflection on nouns and determiners via an offline picture selection task. The results show that morphological salience overrides suffix ambiguity, and that syncretic suffixes undermine the identification of the gender value. Receptive vocabulary proficiency and cumulative length of exposure to the heritage language were found to affect gender comprehension. Study 2 (Chapter 3) examines the comprehension of case on nouns and determiners via an offline truth value judgement task. The results demonstrate that ambiguous suffixes do not override paradigmatic syncretism in structures with non-canonical word order and that Greek heritage children had lower accuracy than the monolingual peers but still higher accuracy than other reported Greek heritage children in the US. Gender comprehension accuracy from Study 1, receptive morphosyntactic proficiency, receptive vocabulary proficiency as well as cumulative length of exposure to the heritage language influenced case comprehension accuracy. Study 3 (Chapter 4) investigates the comprehension of referential short distance which-questions via an offline visual world paradigm task. The results reveal that non-ambiguous case suffixes when presented early in the sentence do not modulate comprehension and that ambiguous case suffixes in structures with non-canonical word order are not interpreted using the case-marking cues of the heritage language. The combination of gender comprehension accuracy to the heritage language from Study 1, case comprehension accuracy to the heritage language from Study 2 as well as receptive vocabulary proficiency to the dominant L2 contributed to heritage children’s sentence interpretation accuracy and strategies.

In the Discussion (Chapter 5), we argue that these findings taken together as a whole suggest that the heritage language has quantitative differences with the baseline control grammar found in native speakers of the heritage language. Heritage language children seem to use sentence comprehension strategies such as SVO word order over non-canonical sentences from their dominant L2 in structures of their heritage language that present variability and as such are vulnerable. The variation shown in sentence comprehension strategies provides evidence for future research that can explore how the sentence comprehension strategies of heritage language children unfold, and how factors like quality and quantity of the attrited parental input affect the development of the heritage language.