Abstract
This thesis investigates syllabification, reduplication, and the interaction between syllabification and reduplication in the Amis language.Considering syllabification, how segments are syllabified and the distribution of vowel /?/ in Amis is illustrated first. Then two competitive solutions – the epenthesis analysis and the deletion analysis -- are offered to account for the presence/absence of inter-consonantal vowel /?/ which affects the syllabification. In the epenthesis considerations, vowel /?/ is inserted between consonant clusters which do not violate coda condition. Contrarily, in the deletion analysis, unstressed vowel /?/ is deleted thus surface consonant clusters are formed. Using Optimality Theory (Prince and Smolensky 1993, McCarthy and Prince 1993a, b), we compare the merits and defects of two analyses in respect of exceptions, stress shift and maximal syllable form; and finally argue for the epenthetic one. Although the deletion analysis is better off in interpreting stress shift, we still favor the epenthesis analysis. The reason is that all accurate forms in different speaking rates can be predicted through the constraint rankings. Moreover, the epenthesis analysis reveals the fact that Amis has surface consonant clusters.Concerning reduplication, the phonological structure of Ca- reduplication and suffixal reduplication are examined. Ca- reduplication copies the leftmost consonant of the base plus a fixed vowel /a/; the vowel /a/ is could be the morphological unmarked vowel (Downing 2000) in the language. Suffixal reduplication mirrors two moras from the right edge minus the final consonant. From the data, we argue that a templatic constraint would better account for suffixal reduplication. Moreover, the Anchoring-C constraint which prevents word-final consonant from copying also plays an important role in suffixal reduplication. Such high-ranked faithfulness constraint in reduplication is discussed, too. Furthermore, the distribution of optional /?/ in reduplicative words displays the interaction between syllabification and reduplication. The high-ranked MaxBR constraint demands the correspondence between base and reduplicant. Therefore, the appearance of epenthetic vowel will be consistent in both base and reduplicant.Finally, we suggest some issues which is worth of further researches: the vowel-glide distinction, the prosodic structure of Amis and a dialectal comparative work of Amis reduplication.