ABSTRACT
Cognitive psychologists studying normal bilingual processing have asked whether L1 and L2 share common or independent systems. The history of cognitive research on whether words in the bilingual's two languages reside in integrated or segregated systems has been reviewed extensively (e.g., Brysbaert, 1998; Chen, 1992; De Groot, 1992b, 1993, 1995; Francis, 1999b; Keatley, 1992; Kroll, 1993; Kroll & De Groot, 1997). To illustrate the way in which claims about bilingual lexical representation have been tested, consider the evidence on cross-language repetition priming. When individuals are asked to make a lexical decision in their L1 (i.e., to decide whether a letter string forms a real word or not), the speed of their responses is typically facilitated if a given stimulus was presented earlier in the experiment (e.g., Scarborough, Cortese, & Scarborough, 1977). In the bilingual version of the paradigm, cross-language translations are substituted; instead of repeating the same word form twice, a word is presented initially and later its translation equivalent appears. For example a Spanish-English bilingual might first see the letter string HOUSE and then the Spanish equivalent CASA in a second phase of the experiment. If words in a bilingual's two languages are represented in a common lexical representation, then priming should be observed across as well as within languages. The results of many cross-language repetition studies failed to obtain evidence for priming (e.g., Brown, Sharma, & Kirsner, 1984; Gerard & Scarborough, 1989; Kirsner, Smith, Lockhart, King, & Jain, 1984; Scarborough, Gerard, & Cortese, 1984). Given the logic of the empirical test, this failure was interpreted as evidence for separate lexicons for each language.
