Local and Global Task Switching Costs in Bilinguals Who Vary in Second Language Proficiency

We examined the relationship between bilinguals' second language (L2) proficiency and their performance on a Stroop switching task, in which a color word (e.g., GREEN) appeared in a congruent ink color (e.g., green) or an incongruent ink color (e.g., red). Participants either read aloud the color word in the word-reading trials or named the ink color in the color-naming trials. Bilinguals who varied in L2 proficiency received 2 pure blocks, consisting of word-reading trials and color-naming trials, respectively, and 1 mixed block, consisting of intermixed word-reading and color-naming trials in an alternating-runs pattern. Comparing performance in nonswitch trials in the mixed block and the pure block provides a measure of global switch costs, whereas differences on switch trials and nonswitch trials in the mixed block reflect local switch costs. Bilinguals with higher L2 proficiency showed a marginally smaller Stroop effect in color naming, a smaller local switch cost in word reading (but not in color naming), and a smaller word-reading versus color-naming task set asymmetry in local switch costs. The latter result was consistent with the language switching finding that the L1/L2 switch cost asymmetry decreased as a function of bilinguals' L2 proficiency. Overall, the current findings support the facilitative role of L2 proficiency in bilinguals' task set switching: Those with higher L2 proficiency have better task set shifting and reconfiguration and updating abilities when they switch from a more difficult task set (color naming) to an easier task set (word reading) in a task-switching paradigm.