Masked priming across languages : an insight into bilingual lexical processing

The masked priming paradigm appears to hold particular promise for the investigation of lexical access in bilinguals because it is possible to test participants in one language (that of the target word) and manipulate the language of the prime without the participants' knowledge that the experiment is about their bilingualism. Despite the apparent elegance of this bilingual paradigm in being able to restrict participants to what has been referred to as "monolingual mode" (see, e.g, Grosjean, 1995), the findings of cross-language masked priming research are still rather varied. In this chapter, we review some key findings and consider the various bilingual lexical models that have been proposed. We argue that in general the data support a two-level account in which lexical form representations are separate from conceptual ones. We suggest that a priming effect can be generated at either of these levels, depending on the nature of the processing task required; and that a joint consideration of task, priming direction, and the types of prime-target relationship may be necessary for a coherent picture to emerge. To illustrate this, we refer to some data from a recent series of across-and within script experiments and attempt to show that a relatively coherent explanation of the data can be given.