Reactive and proactive control in bilingual word production: An investigation of influential factors

Abstract

The present study examined how reactive control (indexed by switching costs) and proactive control (indexed by mixing costs) during bilingual language production was modulated by three factors reflected by different time-courses of stimulus presentation. In three experiments, unbalanced Chinese–English bilinguals named digits in Chinese or English according to a naming cue. In Experiment 1, switching costs reduced when participants had longer preparation time to select the target language to name digits (during the Cue-Stimulus interval, CSI), indicating that longer preparation time helps overcome reactive inhibition. In addition, mixing costs declined drastically at a longer preparation time, indicating that a tiny amount of preparation time allows bilinguals to overcome costs associated with proactively preparing two languages. In Experiment 2, the stimuli were presented prior to the cues, so that participants were given different amounts of time to activate the target lexical nodes in both languages before they were informed of the naming language (during the Stimulus-Cue interval, SCI). Symmetrical switching and mixing costs were observed, indicating that bilinguals can strategically boost activation of the target lexical item in the second language (L2) and attempt to equalize it with its translation equivalent in the native language (L1), when they know previously the specific lexical items to be prepared in two languages. In Experiment 3, different Response-Cue intervals (RCIs) were provided after participants named a digit. It was found that the switching cost asymmetry was more prominent when the time to resolve competition was shorter, while the mixing cost asymmetry emerged only with the longest waiting time. These findings provide the first piece of evidence for the dissipation of the reactive inhibition over time, and suggest that longer preparation would allow the proactive control mechanism to be sensitive the relative proficiency levels of the two languages, leading to stronger proactive control on the dominant language. Taken together, the findings in the present study suggest the dynamic nature of reactive and proactive control in unbalanced bilinguals and have important implications for the current models of bilingual language production, which do not explicitly distinguish the two types of control or address how they adapt to the fine-grained time course of the situation.