Abstract
Interlingual homographs are words from different languages that have identical spelling patterns but unrelated meanings, such as the English word ROOM which is identical in form to the Dutch word for ‘cream’. This paper focusses on the process of response selection in the naming of such homographs, following the activation of the word’s two lexical representations. Not much is known about how the activated lexical representations are mapped onto the appropriate response. One issue is whether participants responding in L2 link the lexical outputs of L1 to the response level in a pure L2 list, i.e., when pure native language words do not appear in the list. The collected data suggest that the answer to this question is ‘no’. Another issue is how is the eventual response is selected in a situation where the lexical access process leads to response competition, i.e., in a mixed language list. The risk of making a wrong-language response was jointly determined by the homograph’s frequency in Dutch and in English. A process of frequency-driven lexical competition and a delay on Dutch responses is necessary but not sufficient to account for the collected mixed data. In addition, a response competition process must be assumed between candidates that are simultaneously activated above a certain threshold.
