Masked Priming by Translation Equivalents in Proficient Bilinguals

Abstract The responses of English-French bilinguals performing semantic categorisation and lexical decision tasks were facilitated by prime stimuli that were non-cognate translation equivalents of the targets (e.g. arbre-tree) when compared to unrelated primes (e.g. balle-tree). These translation priming effects were observed with very brief prime exposures (29–43 msec) and forward and backward masking of the prime. Using the same stimuli, translation priming effects were significantly stronger in the semantic categorisation task then in the lexical decision task. This suggests that the translation priming effect obtained in semantic categorisation is mediated by semantic representations in memory and not the result of form-level connections between translation equivalents, at least for the highly proficient bilinguals tested in the present experiment.