Lexical ambiguity in L2: Homonymy and polysemy among Polish-English bilinguals

The M.A. study reported here aimed at examining the processing of isolated lexically ambiguous words in L2 by fluent Polish-English bilinguals. Just as many people speak more then one language (Cook 2002: 22), most words prove to have more than one meaning or sense (Rodd er al. 2004: 90). Consequently, it appears crucial to add the ambiguity factor to the analysis of the lexicon. Indeed, a variety of semantic theories have tried to address this issue. Problematic as sharp distinctions within lexical ambiguity may be, semantic theories have generally acknowledged the difference between homonymy and polysemy (e.g. Apresjan 1974; Blank 2003). In contrast, psycholinguistic research used to ignore the distinction. This ignorance seems to have led the researchers to misinterpret the cause of the so-called ambiguity advantage in lexical decision tasks (e.g. Rubenstein 1970; Kellas 1988) and stirred much controversy over adequate models of lexical ambiguity processing. According to newer accounts (e.g. Rodd et al. 2002, 2004), the phenomenon is not driven by multiplicity of unrelated meanings (homonymy) but sense-relatedness (polysemy), and metonymy-based polysemy in particular (Klepousniotou – Baum 2007). Since the lexical decision studies conducted so far have focused on lexical ambiguity in L1, the present auditory experiment attempted to verify the more recent findings from a bilingual perspective. Generally modeled on Klepousniotou – Baum (2007), the current project sought to avoid some shortcomings of the original study and added the cross-linguistic factor through cognate homonymous stimuli. Although, again, no homonymy but polysemy advantage was found by the current study, it was metaphorical but not metonymic polysemy which drove the sense-relatedness advantage. Moreover, a significant cognate effect appeared, regardless of the number of meanings. The report tries to account for those findings and suggests improvements for future experiments on the still largely unexplored issue of bilingual lexical ambiguity processing.