Definiteness and Specificity in Second and Third Language Acquisition of Mandarin Chinese

Abstract

This doctoral project addresses critical issues in third language (L3) acquisition, an area less studied than second language (L2) acquisition, by investigating the acquisition of (in)definiteness and (non-)specificity by L2 and L3 Mandarin speakers with prior knowledge of Cantonese and/or English. While English has an article system marking the [±definite] features but not the [±specific] features, Mandarin and Cantonese realise these features through a more complex system of bare nouns, classifiers, numerals, demonstratives, and modifiers. These variations provide unique cases for our empirical study on the acquisition of definiteness and specificity in an article-less language (i.e., Mandarin), a topic that has received less attention in the literature compared to studies on article-based languages. Our participants consist of L1 English L2 Mandarin speakers as well as mirror-image L1 English L2 Cantonese L3 Mandarin Speakers and L1 Cantonese L2 English L3 Mandarin speakers at both early and advanced stages of acquisition, with native Mandarin speakers serving as a control group. They completed both online and offline tasks examining five sets of target properties involving morphosyntactic structures as well as semantic, discourse, and pragmatic conditions. The results for early stages find both facilitative and detrimental transfer from L1 and L2 Cantonese to L3 Mandarin, driven by the typological closeness or structural similarities between these two languages. In later stages, our search for factors influencing acquisition success or failure highlights the distinction between learning and unlearning, the availability of evidence in the input, and the differential acquisition complexity of each target property. Additionally, comparisons between participants’ processing and judgment results for the same target properties reveal discrepancies between their implicit and explicit language knowledge. These discrepancies may be attributed to factors such as learning tasks, cognitive demands, (non-)native processing strategies, and learners’ ability to extract necessary evidence from the input.