Evidence of rapid automatic translation in Korean-English bilinguals using masked implicit priming: An ERP study

Abstract

The present study used a masked implicit priming paradigm to test if L1 to L2 translation occurs automatically and rapidly. Korean-English bilinguals performed a lexical decision task when English L2 targets (e.g., FACE) were translation equivalent to the L1 prime (얼굴 elkwul meaning ‘face’) or had phonological overlap with its translation to varying degrees: moderate (FAKE), minimal (FOOL), or unrelated. The translation equivalent targets resulted in N250 and N400 attenuation, reflecting facilitation in sublexical and lexical mapping of the target words, respectively. Crucially, target words which were phonologically related to the implicitly activated translation equivalent (face–FAKE/FOOL) also demonstrated N250/N400 modulation in the absence of semantic overlap. Additionally, the pattern of effects obtained against the unrelated condition differed between the implicitly related primes, with greater phonological overlap resulting in increased negativity, while minimal overlap led to attenuation. These findings suggest translation via direct lexical association occurring automatically at earlier stages of visual word recognition prior to lexical selection in bilinguals.