Does Sentence Context Constrain Cross-Language Lexical Access?

Open Access
Phelps, Tyler Elisabeth
Graduate Program:
Master of Science
Document Type:
Master Thesis
Date of Defense:
December 18, 2007
Committee Members:
  • Judith Fran Kroll, Thesis Advisor
  • Paola Eulalia Dussias, Thesis Advisor
  • sentence context
  • lexical access
  • cognate
  • bilingualism
Research on cross-language processing at the lexical level suggests that each of the bilingual’s languages is active regardless of the requirement to use one language alone. Words that are language ambiguous (e.g., cognates) appear to be processed differently by bilinguals than monolinguals; constraints such as context, task instruction, and intention do not easily limit access to only one lexicon. However, some reduction in the activation of the unintended language has been found when sentence context is highly constrained semantically. In the present study, we examined the effects of language-specific syntax on the nonselectivity of lexical access for Spanish-English and English-Spanish bilinguals. Specifically, participants named critical target words (cognates or controls) presented in the context of sentences with either language-specific or language-non-specific syntax. We theorized that if the parallel activation of the bilingual’s two languages can be eliminated in the presence of syntactically specific contextual information that corresponds to one language alone, then the cognate effect in word naming, found in previous studies in out-of-context naming, should also be eliminated in context. The results of this study suggest that syntax functions like semantics to constrain activation of the unintended language. We discuss the implications of these results for models of bilingual lexical access and also for code-switching performance.