ABSTRACT
A central issue in theories of bilingual language representation concerns the mapping of form to meaning. Past research on this topic sought to determine whether word forms and concepts in two languages are represented independently or integrated within unitary lexical and conceptual memory systems. Although an indeterminate conclusion concerning this question initially led some to argue that it was not possible to resolve the issue empirically (e.g., Glucksberg, 1984), recent work suggests that the earlier claims of representational indeterminacy failed to take into account distinctions between levels of representation. If one assumes that lexical-level representations are functionally independent for words in two languages (an assumption that seems to be required for languages in which word form is not similar) but that conceptual representations are shared (as they appear to be for words in the first language and pictured objects), it is possible to interpret the earlier contradictory evidence for the two alternative positions (Potter, 1979; Snodgrass, 1984). In general, evidence gathered from tasks that focus on lexical-level processing tends to support the independence view (e.g., Gerard & Scarborough, 1989), whereas evidence gathered from tasks that require semantic-level processing tends to support the integrated view (Durgunoglu & Roediger, 1987). However, although the assumption of two levels that correspond to form and meaning provides an initial framework for thinking about bilingual representation, it does not determine the specific nature of the mappings between form and meaning within or across languages.
