The Organization And Activation Process Of Bilingual Phonological Representations: Evidence From Chinese-English Bilinguals

Most of the previous studies concerning the bilingual mental lexicon focused on the semantic level. The traditional models assume the bilingual mental lexicon is shared at the semantic level but separated at the lexical level (e.g. Kroll & Stewart, 1994). Recent research, however, finds that there exist cross-lingual effects at the lexical level (Gollan & Foster, 1997; Brysbaert, etc. 1999; Dijkstra & van Heuvan, 1998). The BIA+ model (Dijkstra & van Heuvan, 2002) was proposed based on these findings. It claims that the bilingual mental lexicon is shared at the lexical level as well, but the relevant experiments were mostly based on those similar-orthography languages (such as English and Dutch, English and French). The cross-lingual phonological effects may be caused by the coaction of both similar orthography and similar phonology. Whether there is any phonological interaction for those languages which differ greatly in orthography (such as English and Chinese) is still a pending question.Motivated by the current researching situation, this research investigates the organization and activation of the phonological representations in Chinese-English bilinguals. It aims to explore whether any cross-lingual homophone effect can be found between English and Chinese and what their phonological activation time courses are.It has been hypothesized that the phonological representations in Chinese-English bilinguals are shared and the phonological code of logographic Chinese is activated later than that of alphabetic English in visual word recognition, attributed to the different orthographic natures of the two languages. We speculate that it is very likely that the shared phonological representations in Chinese-English bilinguals come into being during the learning process of L2 phonology. Three experiments were designed. Experiment 1 tested the hypothesis of the shared phonological representations and their activation process under the visual modality. It included three sub-experiments under three SOAs (43 milliseconds, 75 milliseconds, and 150 milliseconds), accomplished by a naming task. Experiment 2 was composed of two sub-experiments under the auditory modality, adopting a Chinese lexical decision task and an English lexical decision task, respectively. In Experiment 3 the speakers were required to read sentences aloud as naturally as possible; the native English speakers read sentences in English and the Chinese EFL learners read in both English and Chinese. The formant frequencies of the vowels in their readings were then measured.Experiment 1 corroborates the view of shared phonological representations in Chinese-English bilinguals. It is also confirmed that under the short SOA, the cross-lingual priming effect did not occur when primed by Chinese characters, but this did not hold true when the primes were English. The shared view finds further evidence in Experiment 2 under the auditory modality. The results of acoustic analysis in Experiment 3 verify our postulation that the L2 (the second language) phonological system of Chinese EFL learners was built mostly based on their L1 (the first language) phonological system.The conclusion is that even for the different-orthography bilinguals, phonological representations are integrated in one mental lexicon. With the existing findings on similar-orthography bilinguals, we may infer that except for some language-specific phonological features (such as tone in Chinese), the phonological representations in the bilingual mental lexicon may be shared universally. These findings carry important implications for L2 phonetic teaching and learning.