The impacts of bilingualism upon the human brain, specifically in terms of the neuroplastic and neurofunctional changes upon general cognitive functions and on cognitive decline, are currently attracting much attention not only from the neuroscience and psycholinguistic communities but also from the general public. Researchers are especially focused on the putative protective effects of bilingualism upon cognitive decline.
It is well reported within the literature that bilingualism has an impact upon general cognitive functions. In their Keynote Article in this issue, Baum and Titone (Reference Baum and Titone2014) provide an overarching cross-examination of studies that assessed bilingualism across a variety of populations, investigating the bilingual influence within the domains of linguistic processing and executive control. In moving toward a neuroplastic view on bilingualism, cognitive control, and aging, the authors provide an overview as to why a neurological approach stands to improve the collection of data and the interpretation of results when assessing these domains. Many behavioral studies provide support for positive correlations between bilingualism and the domain of executive control functions; however, as Baum and Titone correctly point out, there are also many studies that counter these findings. The ambiguity associated with behavioral empirical assessments could be better interpreted or strengthened by the neuroanatomical and neurophysiological data gained from neuroimaging techniques that actually measure neural activity in the bilingual brain.
However, unlike Baum and Titone, we believe that structural and functional neuroimaging data concerning the field of bilingualism are not meagerly represented. Abutalebi et al. (Reference Abutalebi, Annoni, Zimine, Pegna, Seghier and Lee-Jahnke2008) review more than a dozen structural neuroimaging papers and more than 40 functional neuroimaging investigations focusing on bilingualism. In addition, research of the past 5 years has produced many exciting and new findings with the help of neuroimaging. What conclusions may we draw from the overwhelming amount of imaging data now available? We believe that by properly interpreting more current and available neuroimaging findings we may already answer some of the several outstanding questions that the authors raise in their concluding remarks. For example, questions regarding which aspects of using more than one language induce neuroplastic changes in the human brain are best addressed by considering the neural basis of bilingual language processing as compared with monolingual language processing. As described in Abutalebi and Green (Reference Abutalebi and Green2007), bilingual language processing is orchestrated by a neurocognitive network that embodies inhibitory control. The theoretical foundations of this neurocognitive model undeniably point toward the presence of a linguistic processing system that differs from that of monolinguals. Neuroplastic changes in the bilingual brain are best expressed as adaptive effects in the neural regions and circuits that mediate specific demands necessary for bilingual language processing. Hence, these neuroplastic changes are more or less absent in the monolingual brain in terms of influencing processes such as executive control, for there is no demand to separate two language systems (Green & Abutalebi, Reference Green and Abutalebi2013).
An account of the bilingual neurocognitive network is schematically described in previous works (Abutalebi & Green, Reference Abutalebi and Green2007, Reference Abutalebi and Green2008) detailing specific neural regions and circuits and their distinct functions and connections within the neurocognitive framework of the inhibitory control model. In brief, we identified the anterior cingulate cortex (ACC) and the presupplementary motor area (pre-SMA) with conflict monitoring between languages and acknowledged a role for the pre-SMA in speech initiation for language switching. The left prefrontal cortex is associated with control of interference, while the inferior parietal lobules are more related to the maintenance of task representations. Subcortical structures, such as the left caudate, are associated with switching between languages (Abutalebi, Della Rosa, Castro Gonzaga, et al., Reference Abutalebi, Della Rosa, Castro Gonzaga, Keim, Costa and Perani2013) and involved in the control of verbal interference (Abutalebi et al., Reference Abutalebi, Annoni, Zimine, Pegna, Seghier and Lee-Jahnke2008; Ali, Green, Kherif, Devlin, & Price, Reference Ali, Green, Kherif, Devlin and Price2010). In our recent update of the model (Green & Abutalebi, Reference Green and Abutalebi2013), we included two further circuits involved in language control: a connection between the prefrontal cortex and right cerebellum (Fabbro, Moretti & Bava, Reference Fabbro, Moretti and Bava2000) and a right inferior frontal cortex subcortical connection to the thalamus (e.g., Aron et al., Reference Aron, Durston, Eagle, Logan, Stinear and Stuphorn2007) and putamen specifically involved in the detection of salient cues.
Granted that bilingual and not monolingual processing intensely engages this circuitry, and considering the demands placed on the execution of two or more language systems, one may come to expect adaptive neural changes; the more the circuitry is engaged the more neural changes we expect to happen in these structures. Structural neuroimaging is a powerful tool with which to measure eventual adaptive neural changes in terms of increased grey or white matter density. For instance, Stein et al., (Reference Stein, Federspiel, Koenig, Wirth, Strik and Wiest2012) recently reported that structural changes in the left inferior frontal gyrus are correlated with an increase in second language (L2) proficiency (for findings in elderly bilinguals, see also Luk, De Sa, & Bialystok, Reference Luk, De Sa and Bialystok2011). Bilingualism, and specifically L2 proficiency, also appear to have neuroplastic effects upon the left inferior parietal lobule (LIPL; Mechelli et al., Reference Mechelli, Crinion, Noppeney, O’Doherty, Ashburner and Frackowiak2004). Furthermore, a grey matter increase in the LIPL is restricted not only to subjects with higher L2 proficiencies but also to those subjects who have been reported to reflect better performance on executive tasks as shown in multilingual children followed-up over a 1-year period (Della Rosa et al., Reference Della Rosa, Videsott, Borsa, Canini, Weekes and Franceschini2013). Likewise, increased grey matter density of the ACC correlates to better conflict monitoring for bilinguals but not for their monolingual counterparts (Abutalebi et al., Reference Abutalebi, Della Rosa, Green, Hernandez, Scifo and Keim2012). In addition, increased grey matter densities in bilinguals as compared to monolinguals were also reported in the basal ganglia, such as the left caudate (Zou, Ding, Abutalebi, Shu, & Peng, Reference Abutalebi, Della Rosa, Green, Hernandez, Scifo and Keim2012) and left putamen (Abutalebi, Della Rosa, Castro Gonzaga, et al., Reference Abutalebi, Della Rosa, Castro Gonzaga, Keim, Costa and Perani2013). These are only some of the structural neuroimaging papers available, but they already provide us with a clear picture: neuroplastic changes take mainly place in the bilingual brain in areas related to executive control (i.e., prefrontal–ACC–LIPL–subcortical). Therefore, these findings further indicate why bilinguals appear to have a cognitive advantage in executive control over monolinguals, which may potentially express itself in a neurocognitive reserve seen in aging bilingual populations.
Another crucial point made by Baum and Titone in their Keynote that tightly relates to our points above is whether cross-language activation generated during in the moment bilingual language processing creates executive control demands that differ from those experienced by monolinguals. Baum and Titone specifically draw attention to the importance of addressing whether such differences are quantitative (i.e., just more of what monolinguals normally experience), qualitative (i.e., fundamentally different from what monolinguals normally experience), or both. We maintain that that the findings accumulated from studies that involve neuroimaging can further aid in linking neuroanatomical interpretations and identifying both qualitative and quantitative differences.
We argue that differences are both quantitative and qualitative between monolingual language processing and bilingual language processing. Differences can be of a qualitative nature when areas in the brain specifically respond to bilingual language processing and remain irresponsive during monolingual processing. Consider the activity seen in the left caudate as an example. In a recent study, we compared multilingual speakers to monolinguals during a language-switching paradigmFootnote
1
and measured specifically the blood oxygen level dependent (BOLD) signal in two distinct areas related to executive control (Abutalebi, Della Rosa, Ding, et al., Reference Abutalebi, Della Rosa, Ding, Weekes, Costa and Green2013). The BOLD signal increased in the caudate specifically for multilingual speakers and decreased when monolinguals completed a switching task within their single language; hence, the monolinguals do not activate that particular brain structure qualitatively. In contrast, the BOLD signal for the pre-SMA/ACC was statistically indifferent during between-language switching for multilinguals and within-language switching for monolinguals; this finding indicates that both groups activate this particular structure for correct response monitoring. It is exactly here where quantitative differences arise: although both groups utilize the ACC to the same degree for everyday action monitoring, bilinguals also need the ACC for language production. Thus, it is the amount of ACC utilization that makes the difference functionally and structurally, as was recently reported for nonverbal conflict resolution (Abutalebi et al., Reference Abutalebi, Della Rosa, Green, Hernandez, Scifo and Keim2012).
We suggest that the bilingual experience strengthens the executive control network, and it is reasonable to believe that bilingual individuals create more connections between the single areas of this network following the rules of Hebbian dynamics. Having more connections is the most plausible interpretation for explaining recent and compelling findings that the bilingual brain is more resistant toward cognitive decline (Bialystok & Craik, Reference Bialystok and Craik2010; Bialystok, Craik, & Freedman, Reference Bialystok, Craik and Freedman2007; Bialystok, Craik, Klein, & Viswanathan, Reference Bialystok, Craik, Klein and Viswanathan2004). Bilingualism as such would accordingly act as a cognitive reserve. We agree with Baum and Titone that future studies should investigate individual differences in order to better describe the contribution of several variables, such as degree of L2 proficiency, amount of exposure to L2, and socioeconomic status. Studies involving behavioral research and neuroimaging techniques must aim to provide comparisons on levels that can access and control individual variability with regard to finding specific correlations between the effects of bilingualism and other cognitive functions. This in turn would provide more insightful data concerning bilingualism and the effects of variables such as aging, paving the way toward a more neuroplastic view of bilingualism and executive control.
The great merit of Baum and Titone's argument is their assertion that the bilingual effect is not accounted for with a simple “yes” or “no” answer and that when interpreting its influence on executive control and aging, bilingualism is better assessed over a continuum that varies both qualitatively and quantitatively. In terms of assessing aging populations, we agree that the use of future neuroimaging studies can prove to be extremely enlightening. Future studies should collect population-level statistics to provide potential neuroanatomical interpretations concerning qualitative and quantitative differences and to provide a possible means of measuring neuroplasticity undergone during longitudinal and experimental conditions concerning bilingualism (i.e., cognitive training).
You have
Access
The impacts of bilingualism upon the human brain, specifically in terms of the neuroplastic and neurofunctional changes upon general cognitive functions and on cognitive decline, are currently attracting much attention not only from the neuroscience and psycholinguistic communities but also from the general public. Researchers are especially focused on the putative protective effects of bilingualism upon cognitive decline.
It is well reported within the literature that bilingualism has an impact upon general cognitive functions. In their Keynote Article in this issue, Baum and Titone (Reference Baum and Titone2014) provide an overarching cross-examination of studies that assessed bilingualism across a variety of populations, investigating the bilingual influence within the domains of linguistic processing and executive control. In moving toward a neuroplastic view on bilingualism, cognitive control, and aging, the authors provide an overview as to why a neurological approach stands to improve the collection of data and the interpretation of results when assessing these domains. Many behavioral studies provide support for positive correlations between bilingualism and the domain of executive control functions; however, as Baum and Titone correctly point out, there are also many studies that counter these findings. The ambiguity associated with behavioral empirical assessments could be better interpreted or strengthened by the neuroanatomical and neurophysiological data gained from neuroimaging techniques that actually measure neural activity in the bilingual brain.
However, unlike Baum and Titone, we believe that structural and functional neuroimaging data concerning the field of bilingualism are not meagerly represented. Abutalebi et al. (Reference Abutalebi, Annoni, Zimine, Pegna, Seghier and Lee-Jahnke2008) review more than a dozen structural neuroimaging papers and more than 40 functional neuroimaging investigations focusing on bilingualism. In addition, research of the past 5 years has produced many exciting and new findings with the help of neuroimaging. What conclusions may we draw from the overwhelming amount of imaging data now available? We believe that by properly interpreting more current and available neuroimaging findings we may already answer some of the several outstanding questions that the authors raise in their concluding remarks. For example, questions regarding which aspects of using more than one language induce neuroplastic changes in the human brain are best addressed by considering the neural basis of bilingual language processing as compared with monolingual language processing. As described in Abutalebi and Green (Reference Abutalebi and Green2007), bilingual language processing is orchestrated by a neurocognitive network that embodies inhibitory control. The theoretical foundations of this neurocognitive model undeniably point toward the presence of a linguistic processing system that differs from that of monolinguals. Neuroplastic changes in the bilingual brain are best expressed as adaptive effects in the neural regions and circuits that mediate specific demands necessary for bilingual language processing. Hence, these neuroplastic changes are more or less absent in the monolingual brain in terms of influencing processes such as executive control, for there is no demand to separate two language systems (Green & Abutalebi, Reference Green and Abutalebi2013).
An account of the bilingual neurocognitive network is schematically described in previous works (Abutalebi & Green, Reference Abutalebi and Green2007, Reference Abutalebi and Green2008) detailing specific neural regions and circuits and their distinct functions and connections within the neurocognitive framework of the inhibitory control model. In brief, we identified the anterior cingulate cortex (ACC) and the presupplementary motor area (pre-SMA) with conflict monitoring between languages and acknowledged a role for the pre-SMA in speech initiation for language switching. The left prefrontal cortex is associated with control of interference, while the inferior parietal lobules are more related to the maintenance of task representations. Subcortical structures, such as the left caudate, are associated with switching between languages (Abutalebi, Della Rosa, Castro Gonzaga, et al., Reference Abutalebi, Della Rosa, Castro Gonzaga, Keim, Costa and Perani2013) and involved in the control of verbal interference (Abutalebi et al., Reference Abutalebi, Annoni, Zimine, Pegna, Seghier and Lee-Jahnke2008; Ali, Green, Kherif, Devlin, & Price, Reference Ali, Green, Kherif, Devlin and Price2010). In our recent update of the model (Green & Abutalebi, Reference Green and Abutalebi2013), we included two further circuits involved in language control: a connection between the prefrontal cortex and right cerebellum (Fabbro, Moretti & Bava, Reference Fabbro, Moretti and Bava2000) and a right inferior frontal cortex subcortical connection to the thalamus (e.g., Aron et al., Reference Aron, Durston, Eagle, Logan, Stinear and Stuphorn2007) and putamen specifically involved in the detection of salient cues.
Granted that bilingual and not monolingual processing intensely engages this circuitry, and considering the demands placed on the execution of two or more language systems, one may come to expect adaptive neural changes; the more the circuitry is engaged the more neural changes we expect to happen in these structures. Structural neuroimaging is a powerful tool with which to measure eventual adaptive neural changes in terms of increased grey or white matter density. For instance, Stein et al., (Reference Stein, Federspiel, Koenig, Wirth, Strik and Wiest2012) recently reported that structural changes in the left inferior frontal gyrus are correlated with an increase in second language (L2) proficiency (for findings in elderly bilinguals, see also Luk, De Sa, & Bialystok, Reference Luk, De Sa and Bialystok2011). Bilingualism, and specifically L2 proficiency, also appear to have neuroplastic effects upon the left inferior parietal lobule (LIPL; Mechelli et al., Reference Mechelli, Crinion, Noppeney, O’Doherty, Ashburner and Frackowiak2004). Furthermore, a grey matter increase in the LIPL is restricted not only to subjects with higher L2 proficiencies but also to those subjects who have been reported to reflect better performance on executive tasks as shown in multilingual children followed-up over a 1-year period (Della Rosa et al., Reference Della Rosa, Videsott, Borsa, Canini, Weekes and Franceschini2013). Likewise, increased grey matter density of the ACC correlates to better conflict monitoring for bilinguals but not for their monolingual counterparts (Abutalebi et al., Reference Abutalebi, Della Rosa, Green, Hernandez, Scifo and Keim2012). In addition, increased grey matter densities in bilinguals as compared to monolinguals were also reported in the basal ganglia, such as the left caudate (Zou, Ding, Abutalebi, Shu, & Peng, Reference Abutalebi, Della Rosa, Green, Hernandez, Scifo and Keim2012) and left putamen (Abutalebi, Della Rosa, Castro Gonzaga, et al., Reference Abutalebi, Della Rosa, Castro Gonzaga, Keim, Costa and Perani2013). These are only some of the structural neuroimaging papers available, but they already provide us with a clear picture: neuroplastic changes take mainly place in the bilingual brain in areas related to executive control (i.e., prefrontal–ACC–LIPL–subcortical). Therefore, these findings further indicate why bilinguals appear to have a cognitive advantage in executive control over monolinguals, which may potentially express itself in a neurocognitive reserve seen in aging bilingual populations.
Another crucial point made by Baum and Titone in their Keynote that tightly relates to our points above is whether cross-language activation generated during in the moment bilingual language processing creates executive control demands that differ from those experienced by monolinguals. Baum and Titone specifically draw attention to the importance of addressing whether such differences are quantitative (i.e., just more of what monolinguals normally experience), qualitative (i.e., fundamentally different from what monolinguals normally experience), or both. We maintain that that the findings accumulated from studies that involve neuroimaging can further aid in linking neuroanatomical interpretations and identifying both qualitative and quantitative differences.
We argue that differences are both quantitative and qualitative between monolingual language processing and bilingual language processing. Differences can be of a qualitative nature when areas in the brain specifically respond to bilingual language processing and remain irresponsive during monolingual processing. Consider the activity seen in the left caudate as an example. In a recent study, we compared multilingual speakers to monolinguals during a language-switching paradigmFootnote 1 and measured specifically the blood oxygen level dependent (BOLD) signal in two distinct areas related to executive control (Abutalebi, Della Rosa, Ding, et al., Reference Abutalebi, Della Rosa, Ding, Weekes, Costa and Green2013). The BOLD signal increased in the caudate specifically for multilingual speakers and decreased when monolinguals completed a switching task within their single language; hence, the monolinguals do not activate that particular brain structure qualitatively. In contrast, the BOLD signal for the pre-SMA/ACC was statistically indifferent during between-language switching for multilinguals and within-language switching for monolinguals; this finding indicates that both groups activate this particular structure for correct response monitoring. It is exactly here where quantitative differences arise: although both groups utilize the ACC to the same degree for everyday action monitoring, bilinguals also need the ACC for language production. Thus, it is the amount of ACC utilization that makes the difference functionally and structurally, as was recently reported for nonverbal conflict resolution (Abutalebi et al., Reference Abutalebi, Della Rosa, Green, Hernandez, Scifo and Keim2012).
We suggest that the bilingual experience strengthens the executive control network, and it is reasonable to believe that bilingual individuals create more connections between the single areas of this network following the rules of Hebbian dynamics. Having more connections is the most plausible interpretation for explaining recent and compelling findings that the bilingual brain is more resistant toward cognitive decline (Bialystok & Craik, Reference Bialystok and Craik2010; Bialystok, Craik, & Freedman, Reference Bialystok, Craik and Freedman2007; Bialystok, Craik, Klein, & Viswanathan, Reference Bialystok, Craik, Klein and Viswanathan2004). Bilingualism as such would accordingly act as a cognitive reserve. We agree with Baum and Titone that future studies should investigate individual differences in order to better describe the contribution of several variables, such as degree of L2 proficiency, amount of exposure to L2, and socioeconomic status. Studies involving behavioral research and neuroimaging techniques must aim to provide comparisons on levels that can access and control individual variability with regard to finding specific correlations between the effects of bilingualism and other cognitive functions. This in turn would provide more insightful data concerning bilingualism and the effects of variables such as aging, paving the way toward a more neuroplastic view of bilingualism and executive control.
The great merit of Baum and Titone's argument is their assertion that the bilingual effect is not accounted for with a simple “yes” or “no” answer and that when interpreting its influence on executive control and aging, bilingualism is better assessed over a continuum that varies both qualitatively and quantitatively. In terms of assessing aging populations, we agree that the use of future neuroimaging studies can prove to be extremely enlightening. Future studies should collect population-level statistics to provide potential neuroanatomical interpretations concerning qualitative and quantitative differences and to provide a possible means of measuring neuroplasticity undergone during longitudinal and experimental conditions concerning bilingualism (i.e., cognitive training).