While considerable work has been invested towards understanding language selection in bilinguals, many questions remain about the nature of these mechanisms. Thus, to speak one of these languages without interference from the other, a bilingual must inhibit non-target language information.

There is substantial evidence that both languages are active when a bilingual is comprehending (e.g., Dijkstra & Van Heuven, 2002 Thierry & Wu, 2007, see Kroll & De Groot, 2009 for review) and producing (see de Bot, 2000 or Kroll & Gollan, 2013 for a review, though see Costa, Pannunzi, Deco, & Pickering, 2017, for a competing account) just one of their two languages.

Speaking two languages seems effortless for many bilinguals despite some of the cognitive challenges that come with it. These results may be explained by a language non-specific selection model in which bilinguals do not restrict selection to one language, with translation facilitation being caused by facilitation at the semantic level offsetting competition at the lexical level. However, language dominance was reversed in mixed-language response blocks, implying inhibition of the dominant language and competition between languages. Both experiments replicated previously reported translation facilitation effects in both single-language and mixed-language-response blocks. Spanish-English bilinguals named pictures in single- or mixed-language-response blocks, with distractors in the target language (Experiment 1), or in the non-target language (Experiment 2).

We investigated translation facilitation effects in PWI with language switching. However, in most previous PWI studies bilinguals responded in just one language, an artificial task restriction. In the picture-word interference (PWI) task, semantically related distractors slow production, while translation-equivalent distractors speed it, possibly implying a language-specific bilingual production system ( Costa, Miozzo & Caramazza, 1999).

Picture language translator