The present study adopted the electroencephalogram (EEG) technique to investigate whether inhibition advantage could modulate different language switches, regardless of the time spent on second language learning.

The inhibitory control (IC) ability of 80 low-proficient Chinese (L1)-English (L2) bilinguals was assessed by the Simon task. Half of these bilinguals were then subdivided into 20 high- and 20 low-IC participants to perform switching between L1 and L2 (L1–L2 switching), and the other half were subdivided into 20 high- and 20 low-IC participants to conduct switching between L1 and Lnew (L1–Lnew switching). All participants were required to name pictures (picture naming task) in their L1 and L2/Lnew in language switching task.

Both response latencies and EEG data were obtained, and then evoked and induced oscillations were calculated using time–frequency analysis.

The results of language switching showed similar naming latencies for L1 and L2/Lnew switch trials in the high-IC group, whereas the low-IC group showed larger naming latencies for L1 switch trials than L2/Lnew switch trials. In contrast, the high-IC group exhibited larger theta evoked and induced power for L2/Lnew switch trials than L1 switch trials at the lexical selection level, whereas the low-IC group did not. These findings indicate that inhibition advantage helps the high-IC group to suppress effectively the non-target word via recruiting bottom-up (evoked oscillation) and top-down (induced oscillation) processes.

The present study was a first attempt to provide evidence that theta oscillation indicates cross-language interference at the lexical selection level.

Inhibition plays a modulatory role in language switching, which is independent of the time spent on second language learning, and such role involves bottom-up (i.e., evoked oscillation) and top-down (i.e., induced oscillation) processes which were mainly evident at the lexical selection level.