Effects of bilingual language experience on basal ganglia computations: A dynamic causal modeling test of the conditional routing model.

Bilingual language control is characterized by the ability to select from amongst competing representations based on the current language in use. According to the Conditional Routing Model (CRM), this feat is underpinned by basal-ganglia signal-routing mechanisms, and may have implications for cognitive flexibility. The current experiment used dynamic causal modeling of fMRI data to compare network-level brain functioning in monolinguals and bilinguals during a task that required productive (semantic decision) and receptive (language) switches. Consistent with the CRM, results showed that: (1) both switch types drove activation in the basal ganglia, (2) bilinguals and monolinguals differed in the strength of influence of dorsolateral prefrontal cortex (DLPFC) on basal ganglia, and (3) differences in bilingual language experience were marginally related to the strength of influence of the switching drives onto basal ganglia. Additionally, a task-by-group interaction was found, suggesting that when bilinguals engaged in language-switching, their task-switching costs were reduced.

Applying math onto mechanisms: mechanistic knowledge is associated with the use of formal mathematical strategies.

It is notoriously difficult for people to adaptively apply formal mathematical strategies learned in school to real-world contexts, even when they possess the required mathematical skills. The current study explores whether a problem context's mechanism can act as an "embodied analogy" onto which abstract mathematical concepts can be applied, leading to more frequent use of formal mathematical strategies. Participants were asked to program a robot to navigate a maze and to create a navigation strategy that would work for differently sized robots. We compared the strategy complexity of participants with high levels of mechanistic knowledge about the robot against participants with low levels of mechanistic knowledge about the robot. Mechanistic knowledge was significantly associated with the frequency and complexity of the mathematical strategies used by participants, suggesting that learning to recognize a problem context's mechanism may promote independent mathematical problem solving in applied contexts.