The effects of bilingualism on conflict resolution: Evidence from an ERP study with Greek-Albanian bilinguals.

Abstract

Event Abstract Back to Event The effects of bilingualism on conflict resolution: Evidence from an ERP study with Greek-Albanian bilinguals. Vasiliki Salvari1*, Christos Frantzidis2*, Panagiotis D. Bamidis2*, Elisavet Chrysochoou1* and Ana B. Vivas1* 1 The University of Sheffield International faculty, City College, Psychology, Greece 2 Aristotle University, Greece A bilingual advantage in non-linguistic cognitive control tasks has been frequently reported in adults [1]. This benefit has been explained by suggesting that the control mechanisms used for effective switching between languages are also deployed in tasks requiring the flexible control of attention. Several researchers, however, have questioned the bilingual advantage [1, 2, 3], suggesting it might lie in uncontrolled factors that can influence cognitive performance (e.g. Socio-Economic Status / SES), or it might be subject to the characteristics of the samples and tasks employed in each study. In the light of a large number of studies involving behavioral measures [often offering controversial evidence, see discussions in 5, 6, 4, 3] but of the limited in number studies with brain-based measures [11], this project aimed at providing a clearer insight, by exploring the brain mechanisms underlying the suggested conflict resolution advantage in both younger and older Greek-Albanian bilinguals of lower SES level, who were strictly matched with Greek-speaking monolinguals on age, gender, non-verbal intelligence, and SES (a factor often ignored in relevant studies; see Morton &Harper, 2007) . An extended language use questionnaire, vocabulary tests in both languages, as well as a language-switching computerized task [7, 4], were also used in an attempt to determine bilingualism type (i.e. early versus late, balanced in both languages versus dominant in one language bilinguals) and allow a more thorough discussion of the findings. . Neurophysiological data acquisition was performed through a Nihon-Kohden device equipped with 64 active electrodes; 57 of them were used for measuring brain signals, one as ground electrode placed on an anterior position, and two as reference electrodes placed behind the ears. There were also eight trigger channels and four electrodes for recording the electrooculogrammic (EOG) movements. During the examination participants were asked to sit in a comfortable chair. Firstly, resting-state data acquisition was employed in a three-minute eyes open and then five-minute eyes closed task. These recordings were used to extract resting-state oscillatory activations and features related with the default-mode network activation. Then, computerized versions of the Simon and Stroop tasks were employed to engage participants in conflict resolution assessments. Specifically, in the context of a numerical Stroop task, participants were each time presented with two digits on the screen (a smaller and a bigger one) and they are requested to indicate the one that is physically larger, while ignoring the numerical magnitude of the digits. In each trial, the numerical size difference is either in the same or in the opposite direction of the physical size difference of the digits (congruent or incongruent trials, respectively); for neutral trials, digits with the same numerical value are used. In the context of a Simon task [9, 10, 2, 8], participants were presented with either incongruent or congruent trials (i.e. where the location of the cue presented each time on the screen was supposed to be responded to with a key in the opposite or the same side of the keyboard, respectively). Working demands were also manipulated in the Simon task (low versus high demands), by relating one or two cue types to each response key, in separate relevant blocks. The electrophysiological analysis focuses on specific ERPs measurements which were previously demonstrated to differ among monolinguals and bilinguals in terms of conflict monitoring, resource allocation, categorization of stimuli and error-related stimulus processing [11]. Therefore, we solve the inverse problem to extract cortical activations which are computed through the Brainstorm graphical user interface [12]. Specifically, the analysis focuses on N2, P3 and error-related negativity (ERN) components. The N2 component peaks 200-350ms following a stimulus and it is assumed to be involved in conflict monitoring; that is, an increased conflict monitoring capacity (demonstrated in several studies employing behavioural, rather than brain-based measures) would be expected to be associated with a larger N2 amplitude. Bilinguals were also expected to show smaller delays in P3 latency for incongruent trials than monolinguals [10], indicating longer stimulus categorization time. Finally, bilinguals were expected to experience greater post-response conflict than monolinguals in the case of errors, due to their assumed greater performance monitoring [11] this should result to larger ERN amplitudes for the bilinguals participants [13] Data collection is on-going. The above-mentioned functional brain-related data is compared to the pattern of the results obtained by the behavioral measures of conflict resolution capacity in both language and ages groups (i.e. reaction times and accuracy rates in the Stroop and Simon tasks), in an attempt to discuss whether any observed differences in brain responses actually represent advantages for bilinguals (in contrast, for example, brain-based differences were not accompanied by behavioral performance differences in the Kousaie & Phillips, 2012 study). The findings are overall discussed in relation to the suggested bilingual advantage in cognitive control, taking, however, into account the characteristics of the present sample (i.e. SES level, age group, etc).