Modeling the control of phonological encoding in bilingual speakers