Martinez, Francisco Emigdio (2005-02). Exploring figurative language processing in bilinguals: the metaphor interference effect. Master's Thesis. Thesis uri icon


  • While studies suggest that figurative, or non-literal, meanings are automatically activated in single language users, little is known about how language proficiency may influence the automaticity of non-literal meaning activation. The present research sought to address this issue by comparing figurative language activation in Spanish-English bilinguals. An interference paradigm (Glucksberg, Gildea & Bookin, 1982) was used in which participants were to judge the literal truth or falsity of statements of the form Some Xs are Ys. Judgments on this task are typically slower to statements that, though literally false, are metaphorically true (e.g., Some lawyers are sharks), suggesting that metaphorical meanings are non-optionally activated (at least in single language users). The present research involved four experiments: Experiment 1 conducted with English-speaking monolinguals, replicated the metaphor interference effect; in Experiment 2 the effect was replicated in English-dominant and in balanced bilinguals tested only in English. Experiment 3 conducted with bilinguals tested in both languages, showed that the metaphor interference effect was not obtained in either language in English-dominant bilinguals and was obtained in Spanish only in the balanced group. The findings from Experiments 1 and 2 support the view that nonliteral (metaphoric) meanings are automatically accessed in monolinguals and bilinguals alike. Experiment 3 involved a fewer number of metaphor trials per language, raising the possibility that this procedural difference may have led to a weakening of the metaphor interference effect. This possibility was directly tested in Experiment 4, conducted with English-speaking monolinguals presented with the same number of metaphor trials as the bilinguals in Experiment 3. The results showed a clear metaphor interference, even with the reduced number of trials. As such, the findings of Experiment 3, where a metaphor interference effect was obtained only for Spanish items, are somewhat equivocal: at face value, they suggest that the effect is modulated by language proficiency. Alternatively, the metaphor interference effect may turn out to be present in both languages, but may simply have been obscured by variability owing to the small sample size per language order. Which of these two interpretations turns out to be valid will depend on additional testing. Implications of the present findings for theories of the organization of the bilingual representational system are addressed.

publication date

  • February 2005