| This dissertation investigates the timing of consonant clusters in various first languages (L1s) and in second language (L2) English. Previous studies on L1 consonant timing have shown that some languages exhibit more consonant overlap than others. However, there has been no systematic comparison among different L1s, and only one study has examined consonant timing in both L1 and L2. The goal of this dissertation is to provide a systematic cross-linguistic analysis on consonant timing in various L1s, and investigate how L1 timing may affect the acquisition of L2.;Consonant clusters in word-initial, word-medial, and across-word positions in four L1s (English, German, Cantonese, and Japanese; selected by the pilot experiment) and L2 English are studied using EMMA (Electromagnetic Midsagittal Articulometer). It is hypothesized that there are cross-linguistic differences in L1 consonant timing, and L1-influenced differences in L2 English timing.;In L1, the results show cross-linguistic differences in the amount of lag between consonants: Cantonese is the least lagged; Japanese is the most lagged; English and German are less lagged than Japanese but more lagged than Cantonese. In the majority of L1s, across-word clusters are more lagged and more variable than word-medial clusters. Within the framework of a recently developed coupled oscillator model of syllable structure, a coupling structure is posited for each L1 that can account for these differences and similarities. The results in L2 English show that all L2 speakers on average are able to largely reproduce the English pattern. Further examination on individual speakers reveals cross-speaker variation in the achievement of the English timing pattern that seems to closely correlate with the amount of time spent in the US. L1-influenced difference is found only in inter-gestural variability in across-word clusters, suggesting that the English coupling structure is acquired more readily in word-medial clusters in L2.;Additionally, following exemplar theory, the effect of word frequency is examined in native and L2 English. Although both native and L2 English exhibit the predicted frequency effect, they differ in the realization of the effect, suggesting that L2 speakers do not reproduce the native English effect, but rather produce their own frequency effect. |
