How speaking two languages affects executive functions has been a longstanding debate and the mechanisms underlying the observed cognitive advantages of bilingualism remain unspecified. Here, using multivariate pattern classification methods, we decoded spatial patterns of neural signals associated with Flanker task performance in mono-dialectal and bi-dialectal speakers of Chinese. While univariate approach to even-related potentials (ERPs) showed no between-group difference, decoding accuracy of ERPs was reduced in bidialectal as compared to mono-dialectal speakers in both congruent-neutral and incongruent-neutral classifications. There was no effect of bidialectalism, however, on decoding accuracy of alpha-band oscillations, an electrophysiological index implicated in inhibition. Behavioural data analysed using the Drift Diffusion Model (DDM) showed facilitating effects of bidialectalism on non-decision times but no effect on drift rates. These findings demonstrate that using two dialects on a daily basis enhances general attentional deployment rather than affecting specific component of executive functions such as inhibitory control. Given that the two dialects of Chinese differed almost exclusively in phonology, the bidialectalism effect was most likely motivated by resolving phonological competition at lexical processing level.
| INTRODUCTIONResearch has shown that bilingual individuals automatically activate both languages even when functioning in a single-language language context (e.g., Jacobs et al., 2016;Thierry & Wu, 2007). The target language is selected by inhibiting lexical representations of the non-target language (Green, 1998;Green & Abutalebi, 2013). This unique language experience of bilinguals has been associated with enhanced domain-general cognitive abilities such as executive functions (Bialystok, 2017(Bialystok, , 2021Bialystok et al., 2012). Executive function is an umbrella term for higher-order cognitive skills that include, but not limited to, suppression of interference from irrelevant Abbreviations: DDM, drift diffusion model; ECOC, error-correcting output codes; EEG, electroencephalography; ERP, event-related potentials; HDDM, hierarchical drift diffusion model; MCMC, Markov chain Monte Carlo; MVPA, multivariate pattern analysis; SVM, support vector machine.