Universal and language-specific processing : the case of prosody

A key question in the science of language is how speech processing can be influenced by both language-universal and language-specific mechanisms (Cutler, Klein, & Levinson, 2005). My graduate research aimed to address this question by adopting a crosslanguage approach to compare languages with different phonological systems. Of all components of linguistic structure, prosody is often considered to be one of the most language-specific dimensions of speech. This can have significant implications for our understanding of language use, because much of speech processing is specifically tailored to the structure and requirements of the native language. However, it is still unclear whether prosody may also play a universal role across languages, and very little comparative attempts have been made to explore this possibility. In this thesis, I examined both the production and perception of prosodic cues to prominence and phrasing in native speakers of English and Mandarin Chinese. In focus production, our research revealed that English and Mandarin speakers were alike in how they used prosody to encode prominence, but there were also systematic language-specific differences in the exact degree to which they enhanced the different prosodic cues (Chapter 2). This, however, was not the case in focus perception, where English and Mandarin listeners were alike in the degree to which they used prosody to predict upcoming prominence, even though the precise cues in the preceding prosody could differ (Chapter 3). Further experiments examining prosodic focus prediction in the speech of different talkers have demonstrated functional cue equivalence in prosodic focus detection (Chapter 4). Likewise, our experiments have also revealed both crosslanguage similarities and differences in the production and perception of juncture cues (Chapter 5). Overall, prosodic processing is the result of a complex but subtle interplay of universal and language-specific structure