T3 Sandhi: A Study Of Sandhi Domain

As a highlighted case of tone sandhi, T3 sandhi in Mandarin Chinese has a history of being concerned by linguistic researchers. As is known, the phonological rules of Mandarin T3 sandhi have been generally acknowledged and a bound of researches for this realm have come into being. Most of them, however, are about the linear description of it and relatively fewer concern foot, stress and intonation, which closely relate tone sandhi in Chinese. In view of the study situation about tone sandhi, this thesis, as an attempt, analyzes sandhi domain of Mandarin T3 sandhi to find out the hierarchically deep behaviors that govern T3 sandhi and then to explored how phonological processes work. Ultimately nonlinear phonological patterns of T3 sandhi are put to the surface. Likewise, the study provides the study of Chinese sandhi with a new attempt. It suggests that when studying Chinese sandhi, sandhi domain should be taken into account.To make the study reliable, this thesis adopts quantitative approach for doing the study and all example sentences ,phrases and words are chosen from the contemporary standard publications of the main land of China.In order to provide a plausible support to the argument, the thesis, in the beginning, reviews the relevant phonological theories on sanhi and tone sandhi, which cover from the standard generative theory(linear theory) to nonlinear phonological theories, namely, Autosegmental Phonology Theory, Prosodic Phonology Theory as well as Optimality Theory. Afterwards it states T3 sandhi in Mandarin Chinese(or Beijing Mandarin) to elicit the vital conceptual terminology minimal rhythmic unit(MRU). It is constructed in a two-pass fashion and it is IP(intonation phrase)-bound: that is, an utterance is first broken into IPs, then within IPs syllables are grouped into MRUs. MRUs do not cross IP boundaries. When it comes to bound sand domain of T3 dandhi, according to the main trend of disyllablizaton of modern Mandarin words, sandhi domain is bounded by a set of hierarchical constraints. Meanwhile, the word formation and the sandhi of clitics are both discussed. At last, the thesis makes a simple conclusion: Through studying dandhi domain of T3 sandhi, this thesis makes it to analyze the phonological patterns of T3 sandhi scientifically. S-structure is organized into prosodic units called MRUs. The resulting prosodic structures then serve as input to T3 sandhi(TS).MRU construction is stratum-ordered: it applies(cyclically) at the lexical level, then iteratively across the board at the phrasal level. Likewise, TS applies first within the MRU, then across MRUs. It also points the implication and the limitation of the study.