How do bilinguals initiate language control?

During bilingual language processing, words in the intended language and the non-target language are activated in parallel, which leads to cross-language interference. For example, when producing a word in one language (e.g., “dog” for a Dutch-English bilingual), evidence has shown that the translation equivalent in the other language (“hond”) is also activated. To resolve this cross-language interference, a process known as language control is implemented. Since this process is necessary for bilinguals to successfully process language, it has received substantial theoretical and empirical attention. However, many basic questions regarding language control remain unresolved. In the current project, I propose to investigate one of these open questions, namely how language control is initiated. More specifically, I will investigate two contradictory proposals of how language control is initiated. One proposal is that language control is initiated by activation of the intended language, which results in inhibition of word representations in all other languages. The other proposal is that a high level of cross-language interference initiates language control.