Issues in Chinese Prosody : Conceptual Foundations of a Linguistically-Motivated Text-to-Speech System for Mandarin

Issues in Chinese prosody: conceptual foundations of a linguistically-motivated text-to-speech system for Mandarin Richard S. Lavin School of Information Science Kyushu Tokai University 9-1-1 Toroku Kumamoto, Japan rick@rslavin.net Abstract I examine various controversial aspects of Chinese prosody–tone structure, syllable structure, stress, and intonation–and stress the need to view all of these as interacting systems, aspects of a hierarchical prosodic structure. I examine various proposals at these various levels of the hierarchy and suggest which are most appropriate. Specifically, I suggest the adoption of Bao's version of syllable and tone, and Chen's account of stress. As for intonation, it is still not possible to make any definitive claims regarding an optimal model, but I examine work done by Kratochvil, Shih, and Garding et al, and suggest promising directions for future work. 1� Introduction I propose to consider here a number of troublesome issues in Chinese prosody. This paper is conceived of as a foundation for a text-to-speech system, though I make no attempt to actually develop such a system at this stage. It is more in the nature of an overview of various approaches that have been taken to date, an attempt to extract broad themes from them, and furthermore an attempt to draw up some kind of "checklist" of features or insights that I feel should be incorporated in any linguistically-motivated model of Chinese prosody. I cast my net fairly widely into the literature, both temporally and in terms of personalities, so as not to discard insights of older or less well-known approaches. An important aim is to develop a model that, while being sufficiently explicit that it could in principal be implemented computationally, makes some kind of sense to phonologists. 259 My main starting point is fairly obvious to many, though controversial in some circles: that we need to incorporate some notion of hierarchical prosodic structure. For the motivation for this decision, I refer readers to work by Anthony Fox dating from the mid-eighties (Fox, 1985, 1986), and summarized in his more recent book (Fox, 2000), which harks back to work done in the 1950's and 60's. Fox looks at prosodic hierarchies proposed by Hockett (1955), Pike (1967), Togeby (1965) and Halliday (1967): it is his claim that, by factoring out differences in terminology and diagramming conventions and by recognizing the difference between primary and secondary hierarchies and between units and features of units, we can say that they were all proposing essentially the same hierarchy. Further, relating known features of Chinese prosody to their domains of operation we can draw up a table as in Table 1. FEATURE UNIT tone syllable stress foot tone sandhi foot or intermediate phrase intonation intonational phrase Table : Units and features of units This is not especially novel, nor is it entirely uncontroversial, but it will serve as a starting point. What I propose to do is go through these levels and features in turn and survey some of the main proposals that have been made. This will of course preclude any …

Content truncated to 500 words