Acha, J. 1 , Laka, I. 2 & Carreiras, M. 1
1 Basque Center on Cognition, Brain and Language
2 University of the basque Country
One of the major challenges for researchers in visual word recognition is to understand how word representations are organized and retrieved in the bilingual mental lexicon. Using a single presentation lexical decision and a word naming task, we examined how Spanish-Basque bilinguals identified Basque and Spanish cognates (equal form and meaning) and false friends (equal form and different meaning), compared to their respective controls (translation words that had a different form for cognates and words differing in meaning and form for false friends). For each word type, we manipulated the orthographic and phonological overlap. Thus, a Spanish word could be spelled the same as a Basque word (total overlap: bata-bata), could be phonologically equal but orthographically different (phonological overlap condition: coma-koma), or orthographically equal but phonologically different (orthographic overlap condition: plaza-plaza). In both tasks cognates were easier to process than false friends compared to their controls, and words with total overlap were processed faster that words in the phonological or orthographic overlap condition, showing a facilitative effect of cross-linguistic orthographic and semantic similarity. However, in the lexical decision task, the phonological overlap condition led to larger latencies and higher number or errors than in the other two conditions. In the naming task, this pattern was observed in the orthographic overlap condition. This was particularly so when words where identified or named in Basque (participant´s L2). Our results show that, beyond the key role played by semantics and language dominance, orthographic and phonological properties of words exert a different influence on bilingual word recognition depending on the task demand (lexical access or phonological articulation). We interpret these findings in the framework of the bilingual interactive activation plus model of word recognition.