Cross-linguistic influence in second-language acquisition

Abstract

This thesis examines the role of Cross-linguistic influence (CLI) in second-language acquisition, especially in the real-time processing of English inflectional morphology among adult second-language (L2) learners with different first-language (L1) backgrounds. Specifically, the study focuses on two different groups: Vietnamese-English and Norwegian-English learners, whose L1s differ significantly in their morphological systems. To be clearer, English requires morphological marking for subject-verb agreement and tense, while this feature is completely absent in Vietnamese and only partially present in Norwegian.

The main aim is to examine whether learners from these two groups exhibit sensitivity to verb inflectional morphemes in subject-verb agreement (such as third-person singular -s and plural subjects) and the past tense morpheme -ed. The thesis also explores the extent to which CLI affects the accuracy of these learners and the role of proficiency in the accuracy pattern. Data were collected using a self-paced reading (SPR) task combined with grammaticality judgment questions, allowing for the measurement of both accuracy and reaction times during sentence comprehension.

Findings are supported by the theoretical framework of the Full Transfer Potential hypothesis (FTP – Westergaard, 2019) and the Feature Reassembly Hypothesis (FRH – Lardiere, 2007, 2008, and 2009), which suggest that CLI affects L2 acquisition on a property-by-property basis, depending on the structural overlap between L1 and L2. In terms of reaction-time results, Norwegian learners exhibited greater sensitivity to third-person singular and past tense violations, despite showing no accuracy advantage in these conditions—suggesting a cross-linguistic influence on implicit knowledge. Vietnamese learners, by contrast, showed slower reaction times in grammatical plural contexts, possibly reflecting increased cognitive load during subject-verb agreement processing. The findings also demonstrate that higher proficiency levels help mitigate CLI effects in accuracy across grammatical domains. Taken together, the study offers insight into how CLI, proficiency, and grammatical feature complexity jointly shape L2 sentence processing.