Elsevier

NeuroImage

Volume 199, 1 October 2019, Pages 454-465
NeuroImage

Brain network reconfiguration for language and domain-general cognitive control in bilinguals

https://doi.org/10.1016/j.neuroimage.2019.06.022Get rights and content

Abstract

For bilinguals, language control is needed for selecting the target language during language production. Numerous studies have examined the neural correlates of language control and shown a close relationship between language control and domain-general cognitive control. However, it remains unknown how these brain regions coordinate with each other when bilinguals exert cognitive control over linguistic and nonlinguistic representations. We addressed this gap using an extended unified structural equation modeling (euSEM) approach. Sixty-five Chinese-English bilinguals performed language switching and nonverbal switching tasks during functional magnetic resonance imaging (fMRI) scanning. The results showed that language control was served by a cooperative brain network, including the frontal lobe, the parietal cortex, subcortical areas, and the cerebellum. More importantly, we found that language control recruited more subcortical areas and connections from frontal to subcortical areas compared with domain-general cognitive control, demonstrating a reconfigurable brain network. In addition, the reconfiguration efficiency of the brain network was mainly determined by general cognitive ability but was also mediated by second language (L2) proficiency. These findings provide the first data-driven connectivity model that specifies the brain network for language control in bilinguals and also shed light on the relationship between language control and domain-general cognitive control.

Introduction

For bilinguals, both languages are activated in parallel (e.g., Guo and Peng, 2006; Kroll et al., 2006; but see Costa et al., 2017) and compete for selection (e.g., Abutalebi and Green, 2007; Hermans et al., 1998) when they speak one of their two languages. The mental process for resolving language competition and selecting the target language is termed language control (Green, 1998).

The most commonly used paradigm to investigate language control in bilingual language production is the language switching task (Meuter and Allport, 1999), which requires bilinguals to name pictures in either their first or second language (L1 or L2) according to cues. The sequence of cues is manipulated to make participants switch between languages, resulting in switch trials (i.e., where a different language is used to name two consecutive trials) and nonswitch trials (i.e., where the same language is used across two consecutive trials). It has been reported that bilinguals respond slower to switch trials than to nonswitch trials, which is termed the switching cost (e.g., Meuter and Allport, 1999). Research has also found an asymmetry in switching costs for unbalanced bilinguals, such that switching from L2 to L1 is more demanding than vice versa (see also Costa and Santesteban, 2004, Exp. 1; Linck et al., 2012; Ma et al., 2016, Exp. 1; but see Christoffels et al., 2007; Declerck et al., 2012). These results have been interpreted as evidence for inhibiting cross-language interference.

In the past decade, many functional magnetic resonance imaging (fMRI) studies have investigated the neural basis of language control. For example, Hernandez and his colleagues (Hernandez et al., 2000) required bilinguals to switch between two languages and found greater activation in the dorsolateral prefrontal cortex (DLPFC). Later studies using similar tasks further revealed that bilingual language control involves cortical-subcortical-cerebellar brain regions (for reviews, see Abutalebi and Green, 2007, 2008, 2016), such as the bilateral prefrontal cortex (Abutalebi and Green, 2008; Branzi et al., 2015; de Bruin et al., 2014; Wang et al., 2009), dorsal anterior cingulate cortex/presupplementary motor area (dACC/pre-SMA) (Abutalebi et al., 2008, 2012, 2013a; Guo et al., 2011), inferior parietal lobe (Guo et al., 2011; Wang et al., 2009), caudate nucleus (Abutalebi et al., 2008, 2013a; Li et al., 2016; Wang et al., 2007), thalamus (Abutalebi et al., 2013a, 2013b) and cerebellum (de Bruin et al., 2014; Filippi et al., 2011; Guo et al., 2011; Wang et al., 2009). These brain areas enable bilingual speakers to successfully select a target language.

It is thus reasonable to assume that all the brain areas mentioned above do not work separately, but in a collaborative way. As proposed by Abutalebi and Green (2007, 2008, 2016), the left prefrontal cortex has connections to the basal ganglia and cerebellum, the dACC/pre-SMA connects to the left prefrontal cortex, basal ganglia and right prefrontal cortex, and the thalamus connects to the right prefrontal cortex and basal ganglia (see Green and Abutalebi, 2013 for details). Nevertheless, the speculated connection diagram based on anatomic or functional connectivity studies still needs further empirical support. Hence, the first question of the present study is how critical brain areas for bilingual language control form a collaborative network.

Furthermore, the above-mentioned brain areas for language control have been frequently reported in studies on domain-general cognitive control (e.g., for task switching, see Dove et al., 2000; for flanker task, see Iannaccone et al., 2015; Zhu et al., 2010; for Simon task see Kerns, 2006; Liu et al., 2004), which is an ability to regulate behavior in accordance with internally defined goals (Braver, 2012; Miller, 2000). Therefore, it is reasonable to speculate that language control may be accomplished by reconfiguring the general mechanism of cognitive control. This speculation is also confirmed by meta-analytic studies that show similar brain areas are engaged in language and cognitive control (see Kim et al., 2012; Luk et al., 2012; Nee et al., 2007; Niendam et al., 2012 for meta-analysis). Several studies have directly compared brain activation of bilinguals during language switching and task switching (i.e., switching between different rules, such as responding to a stimulus according its shape or color) and found highly overlapping brain areas (Blanco-Elorrieta and Pylkkanen, 2016; De Baene et al., 2015; Weissberger et al., 2015). However, it remains unknown how the reconfiguration of bilinguals’ brains can be achieved in a connective way. Thus, our second aim is to reveal how the language control network is reconfigured from the brain network of domain-general cognitive control.

Recently, it has been reported that efficiency in reconfiguring the brain network varies with individual abilities (Schultz and Cole, 2016; Zuo et al., 2018). For example, Schultz and Cole (2016) compared functional connectivity patterns during rest and multiple highly distinct tasks. They took the pattern of similarity to measure how much adjustment the brain needs to meet the demands of a specific task. Less adjustment represents higher reconfiguration efficiency. They found that individuals with higher general cognitive abilities showed higher efficiency in reconfiguring their brain connectivity from resting state to specific task states. However, it is still unknown whether/how individual abilities modulate the efficiency with which the bilingual brain network reconfigures from domain-general control to language control. In addition to general cognitive ability, L2 proficiency might be another potential factor modulating reconfiguration efficiency, due to its important role in bilingual language control (Costa and Santesteban, 2004; Filippi et al., 2014; Videsott et al., 2010). Thus, the third purpose of the present study is to address whether the efficiency, with which the bilingual brain network is reconfigured from domain-general cognitive control into language control, is directly influenced by general cognitive ability and also mediated by L2 proficiency.

Taken together, we aimed to explore three research questions: (a) how critical brain regions cooperate in a connective way for language control in bilingual speakers; (b) how the language control network is reconfigured from domain-general cognitive control; and (c) whether/how the efficiency of brain network reconfiguration is influenced by individual differences in general cognitive ability and L2 proficiency. The answers to these questions will contribute to our understanding of the neural network for bilingual language control and its relationship with domain-general cognitive control.

In the present study, we adopted a language switching task and a nonverbal switching task (Liu et al., 2016) to examine language control and domain-general cognitive control, respectively. It should be mentioned that the nonverbal switching task requires participants to switch between rules when responding to a nonverbal stimulus (see section 2.3.2 for details), which is thought to effectively induce bilinguals’ domain-general cognitive control (Liu et al., 2014, 2016). We recruited Chinese-English bilinguals to perform these two tasks during fMRI scanning. The activation analyses were first conducted to determine the precise brain regions that are critical for language control and domain-general cognitive control. Then, the time course of each critical region was extracted and modeled using an extended unified structural equation modeling approach (euSEM; Gates et al., 2011) for both tasks. EuSEM is a flexible and efficient approach to model the causal interactions of brain regions (i.e., effective connectivity) for cognitive functions, which is based on structural equation modeling (SEM). This approach allows for exploratory data analyses without any prior theoretical assumption and widely applies in numerous studies (e.g., Hillary et al., 2011; Yang et al., 2015; Yang and Li, 2012; Younger et al., 2017). After obtaining the connectivity maps for language control and domain-general cognitive control, we performed paired sample t-tests and complex network analyses to reveal how the brain network reconfigured for domain-general control and language control. We also adopted pattern similarity as an index for brain reconfiguration efficiency and examined the relationship between network similarity and individual abilities, i.e., general cognitive ability and L2 proficiency, via the mediation model (Hayes, 2012).

Section snippets

Participants

Sixty-five Chinese-English bilinguals participated in this study. All of these individuals signed informed consent form before the experiment, which was approved by the Institutional Review Board of Beijing Normal University. All participants were right-handed and had normal or corrected-to-normal vision. None of the participants had color blindness or neurological disorders. Two participants were excluded due to excessive head motion during fMRI data collection (see section 2.5.1).

The

Behavioral results

The response times (RTs) for the language switching task were preprocessed by excluding incorrect trials (4.03%), trials after language incorrect trials (1.52%), absolute outliers (below 300 ms or above 2500 ms, 0.83%) and relative outliers (2.5 standard deviations below or above each individual's mean, 2.40%). The RTs for the nonverbal switching task were preprocessed by excluding incorrect trials (2.97%), trials after incorrect trials (2.82%), absolute outliers (below 200 ms or above 1500 ms,

Discussion

The present study examined the connectivity patterns of the critical brain regions associated with language control and domain-general cognitive control in bilinguals using fMRI. In addition, we further examined how the connectivity pattern was modulated by exerting control on linguistic and nonlinguistic cognitive processing. Finally, mediation analysis was performed to investigate the relationship between the reconfiguration efficiency of the brain network and individual differences in L2

Conclusions

In the present study, we examined the connectivity of brain networks for language and domain-general cognitive control in bilingual speakers. We found that bilingual speakers achieved successful language control by recruiting a well-integrated brain network, which involves the dorsal and ventral frontal lobe, parietal cortex, subcortical areas and cerebellum. Moreover, we found that this brain network for language control is reconfigured from domain-general cognitive control by increasing

Conflicts of interest

No conflicts of interest are declared.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (31871097, 31500924), the Fundamental Research Funds for the Central Universities (2017XTCX04), and the MOE Project of Key Research Institute of Humanities and Social Sciences at Universities (13JJD740009). The first two authors made equal contributions to the study. We thank Dr. Shiqian Zhang and Dr. Eric Pelzl for proofreading the manuscript.

References (129)

  • I.K. Christoffels et al.

    Bilingual language control: an event-related brain potential study

    Brain Res.

    (2007)
  • A. Costa et al.

    Lexical access in bilingual speech production: evidence from language switching in highly proficient bilinguals and L2 learners

    J. Mem. Lang.

    (2004)
  • A. de Bruin et al.

    Domain-general inhibition areas of the brain are involved in language switching: fMRI evidence from trilingual speakers

    Neuroimage

    (2014)
  • P.A. Della Rosa et al.

    A neural interactive location for multilingual talent

    Cortex

    (2013)
  • N.U. Dosenbach et al.

    A core system for the implementation of task sets

    Neuron

    (2006)
  • N.U.F. Dosenbach et al.

    A dual-networks architecture of top-down control

    Trends Cognit. Sci.

    (2008)
  • A. Dove et al.

    Prefrontal cortex activation in task switching: an event-related fMRI study

    Cogn. Brain Res.

    (2000)
  • F. Fabbro et al.

    Language disorders in bilingual patients after thalamic lesions

    J. Neurolinguistics

    (1997)
  • K.J. Friston et al.

    Conjunction revisited

    Neuroimage

    (2005)
  • G. Garbin et al.

    Neural bases of language switching in high and early proficient bilinguals

    Brain Lang.

    (2011)
  • G. Garbin et al.

    Bridging language and attention: brain basis of the impact of bilingualism on cognitive control

    Neuroimage

    (2010)
  • K.M. Gates et al.

    Group search algorithm recovers effective connectivity maps for individuals in homogeneous and heterogeneous samples

    Neuroimage

    (2012)
  • K.M. Gates et al.

    Extended unified SEM approach for modeling event-related fMRI data

    Neuroimage

    (2011)
  • R.H. Grabner et al.

    Superior performance and neural efficiency: the impact of intelligence and expertise

    Brain Res. Bull.

    (2006)
  • T. Guo et al.

    Local and global inhibition in bilingual word production: fMRI evidence from Chinese-English bilinguals

    Neuroimage

    (2011)
  • A.O. Hebb et al.

    The thalamus and language revisited

    Brain Lang.

    (2013)
  • A.E. Hernandez et al.

    Neuroemergentism: a framework for studying cognition and the brain

    J. Neurolinguistics

    (2019)
  • A.E. Hernandez et al.

    Language switching and language representation in Spanish-English bilinguals: an fMRI study

    Neuroimage

    (2001)
  • A.E. Hernandez et al.

    In search of the language switch: an fMRI study of picture naming in Spanish-English bilinguals

    Brain Lang.

    (2000)
  • R. Iannaccone et al.

    Conflict monitoring and error processing: new insights from simultaneous EEG-fMRI

    Neuroimage

    (2015)
  • J.G. Kerns

    Anterior cingulate and prefrontal cortex activity in an FMRI study of trial-to-trial adjustments on the Simon task

    Neuroimage

    (2006)
  • A. Kucyi et al.

    Dynamic functional connectivity of the default mode network tracks daydreaming

    Neuroimage

    (2014)
  • H. Liu et al.

    The effect of domain-general inhibition-related training on language switching: an ERP study

    Cognition

    (2016)
  • X. Liu et al.

    Common and distinct neural substrates of attentional control in an integrated Simon and spatial Stroop task as assessed by event-related fMRI

    Neuroimage

    (2004)
  • F. Ma et al.

    Reactive and proactive control in bilingual word production: an investigation of influential factors

    J. Mem. Lang.

    (2016)
  • V. Menon

    Salience network

  • R.F. Meuter et al.

    Bilingual language switching in naming: Asymmetrical costs of language selection

    J. Mem. Lang.

    (1999)
  • S. Monsell

    Task switching

    Trends Cognit. Sci.

    (2003)
  • T. Nichols et al.

    Valid conjunction inference with the minimum statistic

    Neuroimage

    (2005)
  • T.E. Rohde et al.

    Predicting academic achievement with cognitive ability

    Intelligence

    (2007)
  • M. Rubinov et al.

    Complex network measures of brain connectivity: uses and interpretations

    Neuroimage

    (2010)
  • J. Abutalebi et al.

    Language control and lexical competition in bilinguals: an event-related FMRI study

    Cerebr. Cortex

    (2008)
  • J. Abutalebi et al.

    Bilingualism tunes the anterior cingulate cortex for conflict monitoring

    Cerebr. Cortex

    (2012)
  • J. Abutalebi et al.

    Control mechanisms in bilingual language production: neural evidence from language switching studies

    Lang. Cogn. Process.

    (2008)
  • J. Abutalebi et al.

    Neuroimaging of language control in bilinguals: neural adaptation and reserve

    Biling. Lang. Cognit.

    (2016)
  • A.R. Aron

    From reactive to proactive and selective control: developing a richer model for stopping inappropriate responses

    Biol. Psychiatry

    (2011)
  • A.R. Aron et al.

    Cortical and subcortical contributions to Stop signal response inhibition: role of the subthalamic nucleus

    J. Neurosci.

    (2006)
  • E. Blanco-Elorrieta et al.

    Language switching decomposed through MEG and evidence from bimodal bilinguals

  • E. Blanco-Elorrieta et al.

    bilingual language control in perception versus action: MEG reveals comprehension control mechanisms in anterior cingulate cortex and domain-general control of production in dorsolateral prefrontal cortex

    J. Neurosci.

    (2016)
  • E. Blanco-Elorrieta et al.

    bilingual language switching in the laboratory versus in the wild: the spatiotemporal dynamics of adaptive language control

    J. Neurosci.

    (2017)
  • Cited by (49)

    • Neural circuits underlying language control and modality control in bilinguals: An fMRI study

      2023, Neuropsychologia
      Citation Excerpt :

      Distinct from previous research, we focus on the difference between language and modality control in both production and comprehension conditions. We selected nine seed areas in the gPPI based on the coordinates in Table 2 which have been previously associated with language control (Liu et al., 2020; Wu et al., 2019). The nii.files of the ROIs used in the gPPI analysis were extracted by the MarsBaR tool.

    • Nonverbal cognitive control training increases the efficiency of frontal-subcortical collaboration for bilingual language control

      2022, Neuropsychologia
      Citation Excerpt :

      For example, De Baene et al. (2015) found highly similar activation in the prefrontal cortex, parietal cortex, and subcortical areas when participants performed language switching as well as nonverbal switching tasks. In one of our recent studies (Wu et al., 2019), we further investigated the brain connectivity among the brain regions conjunctively activated across these two tasks. We found a brain network with a similar core-periphery structure and rewired connections from frontal to subcortical regions for language control compared to that for nonverbal cognitive control.

    • “Antagonistic” cooperation of control regions in bilingual language production: An effective connectivity study

      2022, Neuropsychologia
      Citation Excerpt :

      Bilingual language control is the process of selecting the target language and suppressing inferences from nontarget languages during bilingual language processing (Luk et al., 2012; Stasenko et al., 2020; J. Wu et al., 2019).

    View all citing articles on Scopus
    View full text