Empirical approaches assessing adaptations to experience

One way to better assess the adaptations to language experience is in the empirical approaches used in future research. Two directions would ideally be taken. First, multiple measures (both cognitive and neural) would be ideal to more fully assess the range of neurocognitive outcomes. Accompanying this, different statistical approaches will allow us to more holistically assess the numerous relationships between dynamics in language experience and neurocognitive adaptations.

The incorporation of multiple measures can be done in several ways. One approach is to use multiple measures which tap into different cognitive functions (Yamasaki, Stocco & Prat, Reference Yamasaki, Stocco and Prat2018), and include less commonly used measures such as theory of mind tasks (Navarro & Conway, Reference Navarro and Conway2021), which tap into cognitive processes which may also be affected by distinct levels of bilingual experience beyond the ego-driven as discussed in the Systems framework. It is pertinent to note that not all cognitive tasks will necessarily show effects at the level of behavior, and as such the incorporation of neural measures as a complementary method are also encouraged. Including a neural measure (EEG, MRI, etc.) would allow us to see multiple levels of adaptation or effects. Indeed, neural effects of (bilingual) language experience have been shown even in the absence of behavioral effects – both for direct (ego-centric) measures of bilingual language use (Abutalebi, Della Rosa, Green, Hernandez, Scifo, Keim, Cappa & Costa, Reference Abutalebi, Della Rosa, Green, Hernandez, Scifo, Keim, Cappa and Costa2012; Dash, Berroir, Joanette & Ansaldo, Reference Dash, Berroir, Joanette and Ansaldo2019; DeLuca, Rothman, Bialystok & Pliatsikas, Reference DeLuca, Rothman, Bialystok and Pliatsikas2020a; Kałamała, Walther, Zhang, Diaz, Senderecka & Wodniecka, Reference Kałamała, Walther, Zhang, Diaz, Senderecka and Wodniecka2022) and non-direct language exposure (Bice & Kroll, Reference Bice and Kroll2019).

The incorporation of multiple experience and outcome measures also requires more sophisticated statistical modeling to assess the number of potential relationships that exist between these, and how they manifest. Multivariate approaches will be key in unpacking the relationships between diverse language experiences and neurocognitive outcomes. Two examples of such approaches have already been successfully applied within previous bilingualism research. Data-driven approaches such as partial least squares (PLS) are such an example (Anderson, Grundy, Grady, Craik & Bialystok, Reference Anderson, Grundy, Grady, Craik and Bialystok2021; Luk, Anderson, Craik, Grady & Bialystok, Reference Luk, Anderson, Craik, Grady and Bialystok2010). A benefit of PLS is that it is able to take into account several input and outcome variables and derive latent factors that explain meaningful relationships between these variables. To test more theoretically specific relationships and their combined effects, structural equation modeling (SEM) is also an ideal method (Kałamała, Szewczyk, Chuderski, Senderecka & Wodniecka, Reference Kałamała, Szewczyk, Chuderski, Senderecka and Wodniecka2020; Yamasaki & Prat, Reference Yamasaki and Prat2021). With SEM, the key advantage is to directly test theoretical and hypothesized relationships between multiple experiences and outcomes.

It should be noted that no single study would likely have sufficient numbers of participants and measures to test the range of predicted relationships between experience and outcomes. As such, developments in theoretical frameworks linking neurocognitive adaptations to language experiences will also be highly useful.

Neurocognitive models of bilingual adaptation

Methodological developments, such as those described above, would ideally be joined by theoretical approaches that consider the mechanism of adaptation and can also take into account the greater nuance in neurocognition relative to language experience.

A number of models have been proposed which attempt to describe the nature of neurocognitive adaptation to specific aspects of bilingual experience and the associated cognitive demands. Several of these have focused on relating specific experiences (e.g., duration of use, switching, intensity of use, etc.) to neurocognitive adaptations required to meet the associated cognitive demands with these experiences. Duration-based approaches such as the Dynamic Restructuring Model (Pliatsikas, Reference Pliatsikas2020), Bilingual Anterior to Posterior and Subcortical Shift framework (Grundy, Anderson & Bialystok, Reference Grundy, Anderson and Bialystok2017), and Conditional Routing Model (Stocco, Yamasaki, Natalenko & Prat, Reference Stocco, Yamasaki, Natalenko and Prat2014) note a shifting in reliance from frontal to subcortical structures that indicate increased efficiency in handling ongoing cognitive demands from bilingual experience. Alternatively, the Adaptive Control Hypothesis (Abutalebi & Green, Reference Abutalebi and Green2016; Green & Abutalebi, Reference Green and Abutalebi2013) takes an approach of intensity of engagement within specific communicative contexts and the reinforcement of specific networks and circuits commensurate to the degree of engagement. In a similar line of argument, a proposal by Salig and colleagues (Salig, Valdés Kroff, Slevc & Novick, Reference Salig, Valdés Kroff, Slevc and Novick2021) makes a distinction of mental states and traits – with engagement of cognitive resources deployed in a moment-to-moment basis depending on environmental and conversational context. Finally, the Unified Bilingual Experience Trajectories framework (DeLuca, Segaert, Mazaheri & Krott, Reference DeLuca, Segaert, Mazaheri and Krott2020b) has attempted to combine the predictions of several of the above models and extend these predictions across multiple potential neurocognitive outcomes. The notion of the cognitive demands associated with bilingual experience have also been challenged. A recent proposal by Bialystok and Craik (Reference Bialystok and Craik2022) calls for a spectrum approach to conceptualizing the requirements of specific cognitive demands (attentional control) and thus the mechanism by which neurocognitive adaptations would manifest.

On their own, any one of these models or proposals can account for some degree of the variability that is now captured by the Systems framework offered by the authors. However, none of the existing models on bilingual neurocognition either account for the range of language experiences outlined in the systems framework (particularly those beyond language interactions of the individual) and/or make predictions for the range of functional and structural adaptations. Neurocognitive models in the future will thus need to consider and take into account further levels of language dynamics as those outlined in the Systems Framework, but crucially also further outline and make predictions regarding the mechanisms of adaptation.