Evaluating theories of bilingual language control using computational models

Bilingual language control refers to a bilingual’s ability speak exclusively in one without the unintended intruding. It has been debated literature whether bilinguals need an inhibitory mechanism output or non-inhibitory can be used. This paper presents mathematical models instantiating two accounts. The explain how participants’ reaction times production (naming) are impacted by across-trial semantic relatedness and consistency of (same different across trials). models’ predictions were compared data from experiment which participants named semantically-related -unrelated pictures their first second language. Results indicate that within-language facilitation effects abolished after switch, supporting Inhibitory Model. However, was observed over course ‘stay’ trials no switch required, contrary both models. A conducted determine origin this unexpected facilitation, separating spreading activation incremental learning effects. results suggest Experiment 1 due activation. Together, modeling switching abolishes effects, but cumulative interference (created learning) is unaffected switching. suggests (1) non-competitive, (2) between-language competitive (3) plays role bilingual speech production.