Other Language Proficiency Predicts Unique Variance in Verbal Fluency Not Accounted for Directly by Target Language Proficiency: Cross-Language Interference?

The purpose of the study was to investigate cross-language effects in verbal fluency tasks where participants name in English as many exemplars of a target as they can in one minute. A series of multiple regression models were used that employed predictors such as self-rated proficiency in English, self-rated proficiency in a language other than English, a picture naming task used to measure productive vocabulary, the percentage of English use, and the frequency of language switching. The main findings showed that self-rated proficiency in the non-English language accounted for unique variance in verbal fluency that was not accounted for directly by self-rated proficiency in English. This outcome is consistent with cross-language interference, but is also consistent with an account that assumes bilingual disadvantages in verbal fluency and picture naming are due to bilinguals having weaker links between semantic concepts and their phonological form. The present study is also discussed in terms of a broader framework that questions whether domain-general inhibition exists and also whether it plays an important role in bilingual language control.