Abstract/Summary

A growing consensus in bilingual lexical processing research sees the bilingual lexicon as a non-selective system where words from both languages are activated simultaneously during comprehension and production. Less is known about how exactly words are represented and intertwined at different levels and what factors shape the lexicon. The present dissertation investigates the bilingual lexicon’s organization and functioning, employing cross-language visual priming, one of this field’s most productive methodological tools. We assess the contribution of individual-, word- and methodology-level factors. The first study explores the effects of second language (L2) proficiency, L2 use and word frequency in translation masked priming. The second study, employing unmasked primes, focuses on investigating L2 use (L2 proficiency is factored out), word frequency, executive control and degree of semantic overlap. Taking advantage of distributional analyses, the third study compares the cognitive processes recruited during masked and unmasked priming experiments. Overall, our results show that word frequency—prime frequency in particular—is a robust predictor of crosslanguage priming. The contribution of proxies of bilingual experience is more elusive. Some evidence suggests that language use plays a relevant role in the bilingual lexicon’s functioning, while the effect of L2 proficiency in our data is negligible. Further, we show that the degree of semantic overlap and executive control modulate the priming effects. Finally, our results indicate that cognitive recruitment differs in masked and unmasked priming experiments. The findings in this dissertation have several implications: 1) Experiential and word-level factors iii should be further investigated in this type of studies, especially in a continuous manner and focusing on interactions. 2) The models of the bilingual lexicon would benefit from adopting a distributed view of semantic representation. 3) Studies examining response times should include distributional analyses as they provide unique insights into the cognitive mechanisms at play.