| Summary: | 碩士 === 國立中正大學 === 語言學研究所 === 99 === The thesis investigates whether Mandarin speakers’ consonant sonority preferences reflect universal grammatical restrictions, and whether phonological knowledge is independent of phonetic cues. Clements (1990) says that speakers prefer both rising sonority in onset consonant clusters and falling sonority in coda positions. According to Berent, Lennertz, Steriade, & Vaknin (2007), the preference order for marked unattested words of English speaker followed what Clements (1990) argues. Berent et al. (2007) argue that the result reflected the presence of universal grammatical constraints. English allows consonant clusters. Hence, English speakers are familiar with consonant clusters. For this reason, the judgment data from English speakers might be influenced by speakers’ language experience rather than pure universal grammatical restrictions. Mandarin forbids consonant clusters. In order to control the influence of consonant cluster experience, Mandarin participants were divided into two groups in the present study. In one group, participants were familiar with English consonant clusters, and in the other group they were not.
There were three identity judgment tasks in the study. In each experiment, materials for four different sonority sequencing types were designed based on different acceptability from most favored structures (e.g., blif) to least favored structures (e.g., lbif). The auditory stimuli in Experiment 1 were onset consonant clusters. The auditory stimuli in Experiment 2 were coda consonant clusters. Both experiments found that Mandarin speakers preferred consonant cluster types that are attested in English, and disfavored consonant types that are unattested in English. The visual written stimuli in Experiment 3 were onset consonant clusters, and Experiment 3 was designed to investigate the possible influence from phonetic cues. But the results showed a ceiling effect.
To sum up, the present study finds that the Universal Grammar may influence the way how speakers process phonological structures.
|
|---|
