Bilingual word recognition: task demands and the neighborhood density effect

This study investigated the effect task demands have on the processing of cross-language neighborhood density (CND) during word recognition in an English/Spanish bilingual population. Neighborhood Density refers to the total number of orthographically similar words that can arise when a single letter is changed (car; cat/tar). Cross-language ND refers to the total number of orthographically similar words from another language that arises when a single letter is changed (dog; dos/don). Prior research with monolinguals has shown that, during a lexical decision task, higher ND facilitates word recognition in the form of faster reaction time and greater accuracy. However, bilinguals show the opposite effect. That is, increasing the total number of neighbors a word has in both languages leads to slower reaction times and reduced accuracy. A possible explanation for the monolingual/bilingual difference is that cross-language neighbors (those in the language not being identified) require an additional process that expresses itself through lexical access interference. This study investigated how bilingual readers process CND in a Language Specific, lexical decision task (identifying words of one language; LS-LDT) and in a Generalized lexical decision task (identifying words of either language; G-LDT) during mixed language presentation. Reaction time and accuracy for correct identification were measured. Participants were expected to show a CND inhibition effect for words in the LS-LDT, but not for those in the G-LDT. Results showed a general tendency towards supporting the hypothesis but methodological issues limited the scope of interpretation.