The What and Where of Control in Bilingual Language Switching
Slevc, L. Robert
MetadataShow full item record
Speakers of multiple languages must somehow express intended concepts using the appropriate lexical item in the intended language while not producing lexical items from another language that could equally well express the intended concepts. Thus bilingual speakers must presumably manage competition from these items active in their multiple languages in order to successfully communicate. However, it remains unclear where in the process of language production the competition exists, and what mechanisms are used to resolve the competition and successfully produce language. This dissertation set out to more robustly examine the implications of the prominent idea that domain general inhibitory control is used to inhibit the non-target language. To begin, I re-analyzed existing results from studies correlating measures of language switching and inhibitory control using a Bayesian approach. This reanalysis found that much of the previous literature either provides evidence against a relationship between a domain general inhibitory control task and language switching, or finds little to no evidence for such a relationship. Across two experiments, I then assess the role of domain-general inhibitory control in bilingual lexical access using a dual-task design–combining a language switching task with a concurrent task taxing domain-general cognitive control–as well as an individual differences component in the relatively well-powered and pre-registered Experiment 2. In these experiments, I break down the complex process of inhibitory language control into possibly dissociable levels of control (control at the language level and control at the item level) and assess potentially dissociable types of control (proactive control used to bias and monitor for conflict more broadly, and reactive control used for dynamically selecting between languages at a trial by trial level). There was evidence against a role of reactive control in switching between languages at both the language and item level. There was some evidence, however, suggesting a potential role for proactive control or monitoring in a language switching context. Correlations between language switching costs and domain-general measures of inhibitory control suggest that language proficiency and flexibility of control may modulate the ability to reactively control language in a language switching context, however the specificity of these findings demonstrate the complexity of this relationship, in line with the mixed findings in the current literature.