Second language perception of coda morphophonology in spoken English by Mandarin Chinese learners

Acquisition of second language (L2) morphology offers insights into the attainability of L2 structures and differences between the grammatical processing mechanisms of L2 learners and native (L1) speakers. Adult L2 learners often fail to use L2 morphology in a target-like manner, demonstrating inconsistent production of L2 inflectional morphology and insensitivity to L2 morphological errors in comprehension. Learner difficulties are especially acute when the L2 morphological structures are not instantiated in L1 grammar and/or are encoded by phones or phone combinations that are phonologically or phonotactically illicit in the L1. While ample evidence is available on production and reading processing of L2 morphology, less is known about its perception in the spoken L2, and the interface between morphological and phonological influences on L2 acquisition of these structures has not always been explicitly addressed. In this thesis, processing of L2-English word-/syllable-final (coda) morphophonological structures by Mandarin Chinese (ManC) listeners was examined. Their on-line perception of English word-final [n] (e.g., ‘kitten’, ‘fallen’, ‘brighten’), [s] (e.g., ‘house’, ‘box’, ‘books’), and [z] (e.g., ‘breeze’, ‘doors’, ‘trains’) was compared to that of native listeners in a series of self-paced listening and phoneme monitoring experiments. Grammatically incorrect counterparts of target structures that encoded inflectional morphemes were included in each experiment to test L2 listeners’ ability to detect errors (‘The star has just fall’ and ‘The star has falled’; ‘I’ve read many book’; ‘Leave two door open’). Processing of L2-English morphophonology was analysed based on L1-L2 morphophonological congruency, phonotactic context, grammaticality, and utterance position of the target structures; the cognitive and attentional demands of the experimental procedure; as well as the L2 listeners’ length of residence in an L2-speaking environment. Results revealed that L2 processing of morphophonology was determined by L1 phonological rather than morphological influences, with L2 listeners showing more reliable perception of [n] and [s], both of which are available in the ManC segmental inventory, than of [z], which is not present in ManC. Furthermore, L1-L2 phonologically congruent representation of the target structures allowed L2 listeners to successfully attend to L2 morphological complexity and grammaticality, despite the facts that the inflections yielded L1-illegal coda consonant clusters and that the L2 morphemes had no exact equivalents in ManC grammar. L2 listeners demonstrated perceptual sensitivity to the distinctions between polymorphemic versus monomorphemic targets and between grammatical versus ungrammatical inflections. However, their ability to distinguish between morphological variations was more reliable in phoneme monitoring than in self-paced listening, potentially due to the former task’s explicit focus on the target L2 structures. Utterance-final position facilitated L2 processing of L1-L2 phonologically congruent [n] and [s], relative to utterance-medial position. Better utterance comprehension and more sensitivity to errors in self-paced listening experiments were observed for ManC listeners with long L2 exposure than those with relatively short L2 exposure. Overall, the findings of the present research project suggest that L2 learners are able to form mental representations of L2 morphophonology, even when their L1 lacks counterpart morphological structures or has phonotactic restrictions that are violated in the L2. Thus, representational mismatch between L1 and L2 alone cannot account for the variable production and non-native-like reading comprehension of L2 morphophonology reported in previous studies. Instead, it is the processing of L1-L2 incongruent information during time sensitive on-line tasks that is presumed to lead to inconsistent L2 performance in speech production and perception. The ability to efficiently access L2 morphophonology that has been acquired is also constrained by external factors, such as sentential prosody, the nature of the task, and the learners’ experience with L2