Individual Differences in Resting-state Brain Rhythms Uniquely Predict Second Language Learning Rate and Willingness to Communicate in Adults

Abstract

The current study used quantitative electroencephalography (qEEG) to characterize individual differences in neural rhythms at rest and to relate them to fluid reasoning ability, to first language proficiency, and to subsequent second language (L2) learning ability, with the goal of obtaining a better understanding of the neurocognitive bases of L2 aptitude. Mean spectral power, laterality, and coherence metrics were extracted across theta, alpha, beta, and gamma frequency bands obtained from eyes-closed resting-state qEEG data from 41 adults aged 18–34 years. Participants then completed 8 weeks of French training using a virtual language and cultural immersion software. Results replicate and extend previous studies showing that faster learners have higher beta power recorded over right hemisphere (RH) electrode sites, greater laterality (RH − LH/RH + LH) of alpha and beta bands, and greater coherence between RH frontotemporal sites across all frequencies, although only coherence measures survived multiple comparisons. Increased coherence within and between RH networks was also associated with greater posttest declarative memory scores and with more accurate speech during learning. Total speech attempts, in contrast, correlated with bilaterally distributed small-world network configurations, as indexed by lower power and coherence over high-frequency (beta and gamma) bands recorded over frontotemporal networks in both hemispheres. Results from partial correlations and regression analyses suggest that the neural predictors of L2 learning rate, posttest proficiency, and total speech attempts varied in their degree of overlap with qEEG correlates of first language proficiency and fluid reasoning abilities, but that neural predictors alone explained 26–60% of the variance in L2 outcomes.

Relating Natural Language Aptitude to Individual Differences in Learning Programming Languages

“…3: beige) explained an average of 10% of the variance in Python programming outcomes. Similar positive correlations between frontal beta power at rest and learning rate were found in both of our previous rsEEG investigations of natural language learning in adulthood 20,25 . These findings add to the increasing body of literature suggesting that characterizations of resting-state brain networks can be used to understand individual differences in executive functioning 26 and complex skill learning more broadly [19][20][21] .…”

“…This study used a shortened 18-item version of this task, which was developed by splitting the original 36 questions into two, difficulty-matched, subtests based on the data reported in 35 . These parallel forms were previously used in our natural language aptitude research 20,25 .…”

EEG Resting-State and Event-Related Potentials as Markers of Learning Success in Older Adults Following Second Language Training: A Pilot Study

“…Consistent with Prat et al (2016Prat et al ( , 2018, we found a positive relationship between language learning and all assessed frequency bands, which explained 33% of the variance observed. However, only in the beta1 band, power was significantly predicted by L2 development.…”

“…The aim of Experiment 1 was to replicate the studies by Prat et al, (2016Prat et al, ( , 2018 in older adults, and thereby examine whether pre-training resting-state EEG markers can predict L2 aptitude in third age learners following three weeks of L2 instruction only. Using EEG indices measured before the L2 course, we hypothesizedin line with Prat et al (2016) that power in the beta1 band in particular would predict individual L2 development.…”

“…This method allowed us to account for intraclass correlation within subjects and to include all frequency bands within the same model, thereby reducing type I error based on multiple comparisons. The alpha band was used as reference frequency in the model, since we did not expect any significant correlation between alpha and L2 development (Prat et al 2016(Prat et al , 2018. The model did not differentiate between electrodes in order to avoid overfitting and to determine the overall relationship of L2 learning and EEG indices, independent of electrode position.…”