Cross-language neighborhood density effects in early and late bilingual word recognition

Abstract

A central debate in research involving bilinguals is related to how the languages possessed by bilinguals interact while orthographic processing of one language occurs. Past research suggests that there is influence from a bilingual’s non-relevant language when she or he is processing words in the other language. One way of measuring such influence is achieved by varying the number of orthographic neighbors between languages and measuring the difference in reaction times to words with many and fewer cross-language neighbors. In this study, early and late English-Spanish bilinguals, who differed in experiences with their languages, responded to English and Spanish words in a progressive demasking task that differed by the number of orthographic neighbors (many or none) present in the other language. As expected, English words with many cross-language Spanish neighbors were responded to more slowly than English words with no cross-language Spanish neighbors. However, there was no significant difference in reaction times to Spanish words with many or no cross-language neighbors in English, which was unexpected. This pattern was similar in the two groups of bilinguals. Similar results were obtained in a control experiment with monolingual, English-speaking individuals, which suggests that the results obtained from the bilingual study might be due to some uncontrolled lexical variable (e.g., low imageability of specific English words with many Spanish neighbors).