The dissertation aims to better understand the role of native word prosody in the cross-linguistic processing of non-native word prosodic cues (WPCs) (i.e., non-native pitch contrasts and position) at three processing levels: the auditory-acoustic level, the phonological level which taps into abstract representations, and the lexical level where phonological knowledge is encoded to integrate sound-to-meaning associations in word learning. Mandarin, Japanese, and Dutch listeners are selected since these three languages not only differ in their use of WPCs (Mandarin: lexical tones, Japanese: lexical pitch accent, Dutch: lexical stress) but also share commonalities (Mandarin and Japanese: use lexical pitch; Japanese and Dutch: use positional marking cues).
The dissertation applies an AX discrimination task, a sequence recall task and an associated word learning task to investigate the processing of non-native WPCs at the acoustic, phonological and lexical level, respectively, by the three groups of listeners.
The findings show that at the acoustic level, both the acoustic features and the native word prosody influence the perception of non-native WPCs. At the phonological level and lexical level, the use of lexical pitch in the native word prosody plays a pivotal role in the encoding of non-native WPCs. Intonation is found, to some extent, also influence the processing of non-native WPCs at the three levels. Moreover, language-specific and language-general patterns are observed in the processing of non-native WPCs at the three processing levels. Position is found to play different roles in interacting with the processing of non-native pitch contrasts at the three different processing levels.
