Crosslinguistic Influence in L3A: Acquisition of English as a Foreign Language by Arabic, Norwegian and Arabic-Norwegian Adolescents

Abstract

Aims and objectives: This thesis is centered on examining cross-linguistic influence (CLI) in the acquisition of a third language (L3, English) by bilingual Arabic-Norwegian adolescents in comparison with their monolingual controls. The primary objectives of this investigation are to ascertain the origin of CLI, and to determine whether CLI manifests as a property-by-property occurrence or not. Methodology: The participants did an acceptability judgment task (AJT) that assessed six language properties which were chosen based on crosslinguistic similarities and differences between the interacting languages. The properties included were: a) possessive agreement, b) use of the indefinite article, c) use of overt subjects (on these three properties English is similar to Norwegian but different from Arabic), d) word order in non-subject-initial declaratives (SV), e) adverb-verb word order in subject-initial declaratives (Adv-V), and f) subject-verb agreement (SV) (where English is similar to Arabic but different from Norwegian). Data and analysis: A total of 99 participants underwent testing in English, which included an AJT and a proficiency task. The participants comprised 29 L3 learners, 39 L2 learners with L1 Norwegian, and 31 L2 learners with L1 Arabic. Additionally, the L3 learners completed a task in Arabic. Findings: Overall, the findings support the Linguistic Proximity Model (Mykhaylyk et al., 2015; Westergaard et al., 2017; Westergaard, 2019), demonstrating that both Arabic and Norwegian, as the learners' background languages, had an impact on their performance in L3 English. The data analysis of the AJT revealed that the Norwegian control group exhibited the highest level of accuracy, the L3 group fell in an intermediate position, and the Arabic control group demonstrated the lowest level of accuracy. Further findings regarding potential CLI indicated that the Norwegian control group performed significantly better on properties where English and Norwegian behaved the same, as compared to properties where English and Norwegian were different, Similarly, in properties where Arabic patterns with English, the Arabic control group demonstrated a significantly enhanced level of accuracy, in contrast to their performance in properties where facilitative CLI was anticipated to originate from Norwegian. Crucially, however, the performance of the L3 group across the properties was comparable to that of the two control groups. Significance: This study enriches the existing body of research on L3 acquisition and makes a substantial contribution to the ongoing discourse concerning the influence of previously acquired languages in L3A, particularly in relation to the origin and characteristics of cross-linguistic influence (CLI).