Abstract

Recent findings of experimental studies on bilingual word recognition are re-viewed and discussed; moreover, two computational models are presented: a bilingual model of visual word recognition (BIA; Grainger e Dijkstra, 1992; Dijkstra e Van Heuven, 1998; Dijkstra, Van Heuven e Grainger, 1998) and a model of spoken word recognition in bilinguals (BIMOLA; Grosjean, 1988, 1997). Both models make assumptions: a) on the processes involved in bilingual lexical access, b) on the representation of words in two languages. Evidence has accumulated that lexical access may be language-non selective in the first stages of word recognition; however, some methodological and theoretical constraints should be taken into account.