Slide Slam L4
Bilingual speakers’ enhanced monitoring can slow them down
Roksana Markiewicz1, Ali Mazaheri1,2, Andrea Krott1,2; 1School of Psychology, University of Birmingham, UK, 2Centre for Human Brain Health, University of Birmingham, UK
The cognitive demands of bilingualism lead to changes in brain structure and function, particularly in the networks involved in cognitive control (Li et al., 2014; Pliatsikas & Luk, 2016). These changes have been found to lead to behavioural advantages in cognitive control tasks like conflict tasks, but not always (see, e.g., Grundy, 2020). Importantly, cognitive control tasks contain a number of sub-processes. Functional and structural changes might affect these sub-processes in different ways. They might make bilinguals more efficient in some and less efficient in other processes. Performance differences between bilingual and monolingual participants can thus be affected by the balance of various sub-processes. Here we investigated the effect of bilingualism on sub-processes (monitoring and stimulus categorisation) of a conflict task, namely a flanker task (Eriksen & Eriksen, 1974) with congruent (‘>>>>>’) and incongruent (‘>><>>’) arrays of arrows and medium monitoring demand (25% incongruent trials). We examined behavioural and evoked electroencephalographic potentials from a group of young adult bilingual speakers (n=26) and monolingual speakers (n=28). We analysed averaged response times (RTs) and ex-Gaussian analyses of response time distributions. Utilizing an ex-Gaussian analysis allowed to separate a measure of general processing speed (reflected by μ) from a measure of extremeness and frequency of slow responses (reflected by τ). For the evoked potentials we focused on the N2 (implicated to be involved in monitoring) and P300 (implicated to be involved in categorisation) responses. We replicated the flanker congruency effect, both in terms of behavioural and brain responses. In addition, bilinguals had significantly longer response distribution tails (ex-Gaussian τ) compared to monolinguals independent of flanker type, also evident in a trend towards overall slower RTs in bilinguals. Additionally, bilinguals exhibited more pronounced N2 and smaller P3 components compared to their monolingual counterparts, independent of experimental condition, suggesting enhanced bilingual monitoring processes and reduced categorisation effort. Importantly, N2 amplitudes were positively and P3 amplitudes were negatively related to the length of response distribution tails. We postulate that these results reflect an overactive monitoring system (reflected by the more pronounced N2) in bilinguals compared to monolinguals in a task of medium monitoring demand. The enhanced monitoring was followed with fewer resources devoted to conflict resolution and stimulus categorisation (reflected by smaller P3), thus to less effortful categorisation. The monitoring system, however, was rather overactive as it led occasionally to very slow responses. Thus, while the processes of monitoring and categorisation more or less balanced each other out, the less efficient monitoring slightly dominated. These results demonstrate how the efficiency of sub-processes of a task together determine overall behavioural performance and can affect group differences. We propose that the study of the balance of sub-processes in conflict tasks is a fruitful avenue to better understand any functional differences between bilingual and monolingual speakers.