Mandarin Loanword Phonology and Optimality Theory : Evidence from Transliterated American State Names and Typhoon Names

MANDARIN LOANWORD PHONOLOGY AND OPTIMALITY THEORY: EVIDENCE FROM TRANSLITERATED AMERICAN STATE NAMES AND TYPHOON NAMES Hugo Ling-yu Guo Graduate Program in Linguistics National Chengchi University, Taipei, Taiwan, ROC g5555009@grad.cc.nccusedu.tw ABSTRACT This paper aims to examine how the consonant clusters and illicit codas are modified in Mandarin loanwords transliterated from English, and to argue that no rules need to be involved and that a purely constraint-based approach—within the framework of Optimality Theory—can explain the data. The data collected from transliterated American state names and state names display the onset-coda inconsistency in Mandarin loanwords. All the onset consonant clusters in the data are faithfully parsed into Mandarin syllables, with inserting vowels to shun consonant cluster. The coda clusters and illicit codas are generally parsed. However, the coda liquids may be parsed in some cases but unparsed in others. The preference of insertion in Mandarin loanwords can be explained by the interaction between the constraints--*COMPLEX, CODACON, MAX-I0, and DEMO. The distinctive behavior of coda liquids can be accounted for by the effect of MINwD. 1. INTRODUCTION Mandarin transliterated loanwords borrowed from English are prevalent, such as Maryland [mt.n.lond] 4 Malkin [ma.li.lan] 1 . Since Mandarin and English have different syllable structures the latter allows onset and coda clusters and all almost consonants to be codas while the former does not, the syllable structures of the loanwords must be modified or converted while they are borrowed into Mandarin. Three repair strategies—featural change, epenthesis, and deletion—are generally found to operate on these Mandarin loanwords. A question arises: What is the motivation for these strategies? In the previous rule-based studies (Yin, 1984; Chang, 1996: 10-17), several arbitrary rules of deletion and epenthesis are proposed under the condition of context without phonological motivation. These rule- based analyses, as will be presented in section 3.3, also fail to show the connection of related facts. Different from the rule-based perspective, this study will investigate how the consonant clusters and illicit codas are dealt with in transliterated American state names and typhoon names in Mandarin in a constraint-based perspective. 2 It will be argued that the repair strategies on the syllable structures are triggered by high-ranked well-formedness constraints. These constraints interact, in order to increase the well-formedness of the surface forms of the loanwords. It will also be argued that the epenthesis and the disyllabicity effect found in these loanwords result from universal constraints that are present in all grammars, but are masked by the effects of higher-ranked constraints in Mandarin Chinese. The questions addressed are: What constraints are involved in generating the candidate set in Mandarin loanwords? How do these constraints interact to leave one candidate as the optimal output to surface? In what ways is this constraint-based analysis better than the rule-based one in explaining this issue? The Pinyin romanization is used in this study. The pronunciation is transcribed in IPA. The symbol "." indicates syllabification. 2 The term "illicit coda" in this study means the coda in (C)VC. Take, Beth [W], for example. The [0] is not …

Content truncated to 500 words