Cross-language interactions in Chinese-English simultaneous interpreters: Evidence from eye movements and reaction time tasks

The present thesis explores the time-course of cross-language co-activation during simultaneous interpreting (SI) and its underlying cognitive mechanisms. Research on bilingual processing has made the case that both languages are simultaneously active in bilinguals when only one language is being used. The bilingual mental lexicon is thus characterised by strong interactivity across languages as well as across different linguistic dimensions. However, little is known about how languages are activated during SI, a demanding task involving explicit use of both languages and concurrent operation of comprehension and production. In the meantime, interpreting experience has been shown to have an impact on domain-general control functions, a cognitive advantage that is even more profound than that demonstrated in regular, untrained bilingual. It is crucial to understand the cross-language effects experienced by simultaneous interpreters, and more importantly, how the experience shapes the interconnected nature of the bilingual lexicon and the cognitive consequences thereof. The present thesis examines cross- language parallel activation, and the associative link between interpreting-specific parallel activation and domain-general executive function. Professional and student interpreters were recruited to complete (a) a cross-language task (English-to-Chinese simultaneous interpreting) and a within-language task (English-to-English shadowing) and (b) a battery of reaction time tasks indexing executive function. Adopting a visual world paradigm, Study 1 examined the time-course of cross-language co-activation triggered by similarity in word form and meaning during the language production tasks. Evidence of eye movements showed that Chinese translations were activated through both word-form repetition and semantic relatedness, with professionals showing a greater degree of sensitivity to target-language word form during SI task performance, compared to students. Study 2 draws on both on-line eye-tracking data and reaction time task performance. The point-by-point correlation results showed that domain- general cognitive flexibility is associated with the deactivation of target-language word forms during SI. Implications for seminal models of interpreting and bilingual processing are discussed.