The effect of language proficiency,usage,and exposure on cognitive control: A study in early adulthood Chinese learners of English

Abstract

Aims and Objectives/Purpose/Research Questions:

This study investigated the effect of the specific L2 learning experience (i.e., usage, proficiency, and exposure) on cognitive performance in 121 Chinese learners of English.

Design/Methodology/Approach:

The participants were divided into three groups: beginning, intermediate, and advanced learners. They are homogeneous in background variables (e.g., cultural and educational environment) but heterogeneous in the L2 language experience. They performed three non-linguistic cognitive tasks, tapping into multiple dimensions of cognitive functions in visual and auditory domains.

Data and Analysis:

Linear mixed-effect models were applied to RT-based analysis (i.e., analysis based on Reaction Time), and linear regression models were used for accuracy-based analysis. In addition, further multiple stepwise regression analyses were conducted to explore the relationships between individuals’ demographics, language experience, and their performance on the attentional tasks.

Findings/Conclusions:

Three groups were comparable in background measures (e.g., socioeconomic status [SES]) and fundamental cognitive abilities (e.g., working memory and IQ), but differed in specific subcomponents of cognitive functions. Specifically, compared with beginning learners, advanced learners who had a longer length of L2 usage and higher proficiency showed better inhibitory control; intermediate learners who received intensive L2 exposure and had higher proficiency showed better-switching ability and attentional disengagement. Intermediate and advanced learners were comparable in cognitive performance. The results suggest that early adulthood L2 learners experience similar cognitive effects of bilingualism, which are modulated by specific language experience.

Originality:

This is one of the first bilingual studies to incorporate both visual and auditory cognitive functions, while providing additional analyses to investigate the other components of cognitive control.

Significance/Implications:

The results contribute to the understanding of the key aspects of the L2 learning experience that contribute to the emergence of the cognitive effects of young adult learners.

Keywords

Language learner cognitive functions sequential congruency effect early adulthood

Introduction

Learning a second language (L2) is a very common experience in the modern world, as evidenced by the fact that the proportion of bilinguals and multilinguals around the globe is half, if not more, of the world population (Grosjean, 2010). Learning and mastering two or more languages (i.e., bilingualism) has been reported to be associated with some enhancements of cognitive functions, including inhibitory control, switching ability, conflict monitoring, and working memory (Bialystok, 2017; Degirmenci et al., 2022; Grundy, 2020; Ware et al., 2020; Yurtsever et al., 2023). These effects are assumed to stem from the constant management of two jointly activated languages (Thierry & Wu, 2007), especially related to switching between them, selecting the target language, and inhibiting the non-target language (Bialystok et al., 2012; Kroll & Bialystok, 2013). Further neuroimaging evidence indicates a significant overlap of brain regions implicated in language control and domain-general executive control (Green & Abutalebi, 2013; Grundy, Chung-Fat-Yim, et al., 2017). Given the high frequency of young adult L2 learners, it is important to determine whether the cognitive effects associated with bilingualism extend to this particular population.

However, conflicting evidence has been reported across studies, resulting in a debate as to the nature and extent and even the existence of possible effects of bilingualism (Bialystok, 2020; Lehtonen et al., 2018; Paap & Greenberg, 2013). A variety of interacting factors responsible for the complexity and inconsistency of these effects have been discussed, such as the definition of bilingualism, the type of selected tasks, the populations under study, the manner of language acquisition, and types of statistical analyses used among others (Bialystok, 2020). Taking into account the theoretical and methodological aspects of previous research, this study investigates the cognitive effects associated with bilingualism in the context of learning L2.

The role of specific language experience in bilingualism

The dynamic nature of bilingual language processing suggests that multiple factors affect the extent to which dual-language activation impacts language control and in turn influences cognitive control (Bialystok et al., 2012; Kroll et al., 2015). Language proficiency, for example, is one of the core variables when bilinguals are defined (Mishra, 2015), which has been widely investigated at the behavioural level (Luk et al., 2011; Xia et al., 2022), as well as at the neuroanatomical level (Gullifer et al., 2018; Sulpizio et al., 2020). Higher proficiency in L2 is assumed to elicit a stronger interference to the native language (L1) as compared to lower proficiency in L2, leading to increased cognitive demands for language control (e.g., inhibitory control; Hui et al., 2020; Xia et al., 2022). The age of acquisition (AoA) is another crucial factor, which indicates the length of language usage. Studies have suggested that early AoA of L2 (i.e., a longer duration of L2 use) is associated with better inhibitory control (Luk et al., 2011) and attentional monitoring (Kapa & Colombo, 2013). One less-explored aspect of bilingualism is L2 exposure in daily life, which is a proxy for dominance in bilingual language acquisition and development (Unsworth, 2015). Bonfieni et al. (2019), for example, found that more exposure to L2 made it generally easier to access and switch between the first language (L1) and L2, leading to better-switching ability; thus, it modulated language access and control (e.g., switching ability).

Given the complexity and dynamicity of bilingualism, however, it is unclear which facet of language experience (e.g., proficiency, exposure, or usage) would affect which specific aspect of cognitive functions underlying the L2 experience, and to what extent the L2 experience (e.g., when and how the L2 proficiency improves) might contribute to a change in cognitive control. Hence, identifying individuals’ specific L2 experience is crucial for understanding language control in bilinguals; importantly, it may partly explain the reasons behind the inconsistencies in the literature.

The sequential congruency effect

Most studies exploring the effects of bilingualism on cognitive functions have focused on conflict resolution, which refers to the inhibitory control associated with the bilingual experience. An alternative aspect of cognitive control associated with language control could be the disengagement of attention from previous/irrelevant information (i.e., sequential congruency effect [SCE]). The SCE is also referred to as the ‘conflict adaptation effect’ (Botvinick et al., 2001) or the ‘Gratton effect’ (Gratton et al., 1992), which reflects a dynamic, top-down adjustment of inhibitory control in response to the occurrence of interference (Egner, 2007). Grundy, Chung-Fat-Yim, et al. (2017) for instance, reported comparable performance between monolinguals and bilinguals on overall response time and the size of the Flanker effect, but observed significantly smaller SCEs for bilinguals than for monolinguals, indicating faster disengagement of attention from previous trials by bilinguals. More specifically, they divided congruent and incongruent trials into four types of trials based on the previous trial type: two types of current congruent trials (C) following previous congruent trial (cC) and incongruent trial (iC), and two types of current incongruent trials (I) following previous congruent trial (cI) and incongruent trial (iI). This resulted in two types of congruency effects: the c-congruency effect (cI vs. cC) and the i-congruency effect (iI vs. iC). The SCE is indexed by the difference between the two-congruency effects (c-congruency vs. i-congruency effects).

Other studies have interpreted their results as evidence for greater attentional disengagement associated with bilingualism, where high-proficient bilinguals showed an earlier appearance of inhibition of return than that of low-proficient bilinguals (Mishra et al., 2012); the post-conflict slowing (bivalency) effects disappeared early in bilingual children than in monolingual children (Grundy & Keyvani Chahi, 2017). In contrast, other studies have shown the opposite pattern of the effects. Treccani et al. (2009) found a smaller cost of distractor presence (i.e., better inhibitory control) in the condition where the target was not presented in a previously primed position, but this enhanced inhibition turned into a disadvantage in conditions where the previously inhibited information became relevant in the following trial (i.e., negative priming effect). Blumenfeld and Marian (2011) further suggested that this cognitive control might depend on the time interval between inhibition and disengagement of inhibition, and also presumably on differences in the bilingual experience, such as language proficiency.

This study

While many studies have examined the bilingual effects by comparing lifetime monolinguals with bilinguals, fewer studies have investigated whether these effects extend to early adulthood L2 learners. This study aimed to further examine whether the L2 learning experience will benefit cognitive control in three groups of Chinese young adult English learners with varying language experience. More specific research questions are the following:

RQ1. How does the L2 learning experience influence cognitive control?

RQ2. To what extent individuals’ specific language experience can contribute to the cognitive effects associated with bilingualism?

The hypothesis was that a higher level of L2 proficiency and longer length of L2 usage would predict better inhibitory control (i.e., smaller conflict effect in the Attention Network Task (ANT) or Stroop effect in the Stroop task or better performance in the Elevator with Distraction (ED) subtask of the Test of Everyday Attention [TEA]; Hui et al., 2020; Xia et al., 2022), and greater intensity of L2 exposure would predict better-switching ability and faster disengagement of attention (i.e., smaller SCEs in the two computerised tasks; Bonfieni et al., 2019; Costa et al., 2008; Grundy, Chung-Fat-Yim, et al., 2017).

A series of non-linguistic tasks were selected for the following reasons. First, these tasks are sensitive to measuring different aspects of cognitive control (e.g., attention control in the ANT and inhibitory control in the Stroop task) and have been commonly used in previous bilingual studies (Bak et al., 2014; Costa et al., 2008; Hernández et al., 2010; but see Paap et al., 2020 for counterarguments). In addition, language learning entails fundamental and interactive domains, such as the visual domain (e.g., reading and writing) and the auditory domain (e.g., speaking and listening). This suggests that language acquisition and control may demand cognitive resources from both visual and auditory domains. These selected tasks tap into both visual and auditory attention control, thus providing an assessment of cognitive control in the two distinct modalities of language learning (Bak et al., 2016; Vega-Mendoza et al., 2015). Finally, the combination of multiple tasks can avoid the issue of task impurity and the limitations of using a single indicator (Xia et al., 2022).

In addition to traditional standardised analyses of attention indexes in cognitive tasks, we also conducted additional analyses to investigate how language experience influences switching ability (i.e., switching cost; Costa et al., 2008) and attentional disengagement (i.e., SCE; Grundy, Chung-Fat-Yim, et al., 2017). In this study, the switching cost was examined based on the switching conditions that collapsed all trials. In the original study, Costa et al. (2008) examined the switching cost only based on the congruent and incongruent trials. Specifically, ‘switch trials’ collapse all congruent and incongruent trials that involve shifting the mindset between trials, while ‘non-switch trials’ include those congruent and incongruent trials that do not involve switching.

Methods

Participants

A total of 121 Chinese university students who were English learners were recruited from Hubei University, Wuhan, China. Since the contemporary Chinese education system has introduced English learning into primary or secondary education, it is difficult to recruit pure ‘monolinguals’ with no exposure to any other language than their mother tongue. Hence, we recruited a group of participants as a control group who were English learners (i.e., non-English-major) at the beginning level (i.e., the beginning group; n = 61; M _AoA = 9.11 years) with limited proficiency, usage, and exposure to English. During the participants’ recruitment session, they were asked to self-report their overall English proficiency and their ability to hold a conversation in English. Only those who were reported as not functionally fluent in any subcategories of English were recruited into the testing session. By the testing time, they had been enrolled in the university for 2 months and attended two English classes (90 minutes/classes) every week; their usage of English was limited to classroom learning. In the testing session, they were asked to rate their English skills on a language questionnaire adapted from Xia et al. (2022) (the scale is 1–4, marked from ‘poor’ to ‘native/near native’). The proficiency of two participants was too low to be reported and was left blank, 13 participants reported the lowest scores (i.e., 1) for each of the language subcategories, and 22 participants reported at least one subcategory as the lowest score. We compared these participants with the remaining participants who self-rated on language subcomponents equal or higher than 2 on all measures (i.e., background and cognitive measures) and found no group differences in any measures (all p > .05). Thus, we kept the remaining participants in the Beginning group 1.

Two groups of English-major (e.g., Translation and English Education) students who actively use their L2 daily were recruited. The intermediate group (n = 30; M _AoA = 9.23 years) was predominantly first-year (Y1) (i.e., Y1: 73.33%; Y2: 26.67%) undergraduates who had been receiving intensive English-major courses for 2 months and spent significantly longer time on English learning daily (i.e., approximately 6 hours daily and within a 9-hour total course schedule). The advanced group (n = 30; M _AoA = 10.37 years) was mainly Y4 students (i.e., Y4: 67.67%; Y3: 10%; Y1 masters: 23.33%) who had been receiving cumulative similar English learning experience as the intermediate group for at least 3 years, but had much less intensive English exposure and spent fewer hours in English daily. This is because they were approaching the end of their programme and had fewer classes. The main difference between the intermediate and advanced groups was the length of language usage and intensity of language exposure: The advanced group had a longer length of L2 usage, while the intermediate group obtained greater intensity of L2 exposure. Participants’ demographic information is given in Table 1. This study was approved by the Psychology Ethics Committees from both Hubei University and the University of Edinburgh.

Table 1.

Participants’ background information.

	Beginning	Intermediate	Advanced
N/female	61/26	30/27	30/28
Age in years	18.75 (1.29)	18.17 (0.65)	21.47 (1.07)
SES indexed by Parents’ Education ^a	4.96 (1.44)	5.26 (1.58)	4.5 (1.57)
IQ indexed by Raven’s APM ^b	15.43 (3.47)	16.17 (3.41)	15.43 (3.84)
Weekly English learning hours	5.22 (3.59)	42.33 (17.94)	23.02 (11.06)
Corsi tapping task
Forwards	52.15 (9.05)	55 (9.34)	54.17 (8.26)
Backwards	46.21 (9.77)	41.67 (10.93)	44.79 (11.85)
Age of L2 acquisition in years	9.11 (2.44)	9.23 (2.16)	10.37 (2.08)
Self-reported L2 proficiency ^c
Overall	6.98 (2.11)	10.67 (1.81)	11.27 (1.68)
Writing	1.79 (0.61)	2.67 (0.55)	2.86 (0.52)
Reading	1.84 (0.72)	2.8 (0.55)	2.93 (0.52)
Speaking	1.59 (0.59)	2.57 (0.63)	2.7 (0.53)
Listening	1.81 (0.68)	2.7 (0.53)	2.87 (0.51)

SDs are given in parentheses.

Composite score based on the level of parental education. The scale ranged from 1 = primary school; 2 = middle school; 3 = high school; 4 = Bachelor’s or equivalent to; 5 = postgraduate.

Raven’s APM scores were the number of correct items (the total number was 36).

The scale of self-reported L2 proficiency is 1–4, marked by ‘poor’ to ‘native/near native’

Background measures

The Raven’s Advanced Progressive Matrices

The Advanced Progressive Matrices (APM; Raven & Foulds, 1962) were used as a control of nonverbal general intelligence. We adopted Set I (i.e., Item 5 or Item 7) as practice and Set II as the experimental test. Participants were instructed to complete the matrices item by item in 10 minutes and were told that if they were having difficulty with a specific item, they could guess the answer and proceed to the next one. The design of the matrices ensures that the demand of the level gradually increases with the items. Participants started with Item 1 and had to accurately answer as many items as they could. The results were scored as the number of correct items for each participant.

Corsi Tapping Task

The Corsi Tapping Task (CTT; Wechsler Memory Scale-III, Wechsler, 1997) was used as a measure of working memory. This task was presented on a plastic whiteboard (27.5 cm × 21 cm) with 10 blue cube-shaped numbered blocks (3 cm × 3 cm; from 1 to 10). During the testing, the board was placed between the experimenter and the participant, and the numbers were only visible to the experimenter. In the forward condition, the experimenter tapped a sequence of blocks at a rate of approximately 1 second per block in a predetermined order, and participants had to reproduce the tapping number sequence with the same blocks and same order. In the backward condition, all the procedures were the same as in the forward block condition except that the participants had to reproduce the tapping number in the reverse order. There were eight items in each condition, varying from two 2-block trials to two 9-block trials. Accuracy was calculated as the percentage of correct trials.

Background questionnaire

Participants completed a background questionnaire, including both demographic and language-related information. Other factors that have previously been reported to influence cognitive performance in bilingualism research were also collected, including socioeconomic status (SES), musical experience, and video-gaming experience (Bialystok, 2017).

Cognitive tasks

Three non-linguistic cognitive tasks (i.e., the ANT, the Stroop task, and the TEA) were employed to measure different aspects of executive functions in both visual and auditory domains. The computerised tasks were presented with E-Prime (version 2.0) on a 17-inch computer screen. The auditory elevator subtests of the TEA were played through a media player. A schematic representation of each task is depicted in Figure 1.

ANT

The ANT is a well-established assessment of attention control (Fan et al., 2002), which has been widely used in previous bilingualism research (Costa et al., 2008; Xia et al., 2022). Participants were instructed to respond to the central arrow of the five horizontal arrows presented in the middle of the screen either below or above a fixation cross. There were three types of trials: congruent, neutral, and incongruent, and four cueing conditions: single, double, centre and no-cue. Three attentional indices were assessed by the difference in RTs between the following trials: ANT conflict (congruent vs. incongruent trial), ANT alerting (double-cue vs. no-cue), and ANT orienting (centre-cue vs. single-cue). Participants started with a practice block consisting of 24 trials followed by three experimental blocks of 96 trials each. All stimuli were presented randomly at an equal number of times in each block. Feedback on performance was only provided in the practice block.

Number Stroop task

The Number Stroop task adapted from Hernández et al. (2010) was used to avoid any linguistic influence, which is well documented to measure inhibitory control (i.e., Stroop effect). In this task, participants were asked to count digits or symbols presented in the centre of the screen by pressing the keys 1, 2, or 3 on the keyboard while ignoring the numerical value of the digits. There were three experimental conditions: congruent condition, incongruent condition, and neutral condition. Inhibitory control was assessed in this task through the Stroop effect (incongruent vs. congruent trials). Participants were given a practice block with 18 trials which were followed by two experimental blocks of 90 trials each. Feedback on performance was only provided in the practice block.

TEA

The TEA (Robertson et al., 1994) is a well-established clinical assessment of auditory attention, which has been used in previous bilingualism research (Bak et al., 2014; Vega-Mendoza et al., 2015). Since participants’ performance on the elevator with counting usually reached the ceiling level (Xia et al., 2022), we selected the other two subtests of elevator tasks. All tasks were presented through a media player.

a. Elevator with Distraction (ED: 10 trials): Participants were asked to count low tones while ignoring interspersed high tones. Two trials were presented for practice. This task assesses auditory selective attention/inhibition.

b. Elevator with Reversal (ER: 10 trials): Participants were presented with high, middle, and low tones. The middle tones were to be counted while the high and low tones indicated the counting direction (upwards and downwards, respectively). Three trials were presented for practice. This task assesses auditory attentional switching.

Statistical analyses

All analyses were conducted by fitting linear mixed-effect models (LMMs) from the lme4 packages (Bates et al., 2015) into R (Version 3.6.1) unless specified. In the initial analysis, background variables with a continuous scale (e.g., age) were analysed using one-way analysis of variance (ANOVA), while other experience measures on a nominal scale (e.g., gender) were analysed using the chi-square test. Data from the incorrect trials and RTs exceeding 3 standard deviations (SD) of mean were excluded (i.e., 3.10% in the ANT and 5.15% in the Stroop task). In the RT analyses, group and trial type were fixed variables unless specified, and participants and items were random variables. All participants showed comparable performance (both p > .05) and relatively high accuracy rates in the RT-based tasks (i.e., ANT: 98.18% and Stroop: 96.34%); therefore, the accuracy data were not analysed and reported. Linear regression models were used for the analysis of TEA, with the accuracy rate as a dependent variable and group as fixed variables. The Bonferroni corrections were applied with an adjusted significance level of p = .0167. Further multiple stepwise regression analyses were conducted to explore the relationships between individuals’ demographic measures (e.g., sex), language experience (e.g., AoA), and their performance on these attentional tasks. These analyses were conducted through the ‘olsrr’ package. Multicollinearity between predictors was tested using variance inflation factors (VIFs) in each model, and all were below 2. We, in addition, calculated pseudo R² values using the r.squaredGLMM function of the package MuMIn for the LMMs (Barton, 2014), which provides a value for marginal R² (variance explained by fixed effects) and a value for conditional R² (variance explained by both fixed and random effects). We used the r.squaredLR function for the LRs, which provides a value for R² and a value for adjusted R².

Results

Initial analyses

There were no group differences in SES, musical experience, video-gaming experience, IQ, and working memory (all p > .05), indicating comparable fundamental cognitive abilities among groups. As expected, advanced learners were older than both beginning and intermediate learners and beginning learners had a higher proportion of male participants than those of intermediate and advanced (all p < .05). No other group differences were found.

For language-related variables, no group differences were found in AoA (p = .06). Beginning learners had lower self-reported proficiency than both advanced and intermediate learners (all p < .05), with no group differences between later two groups (all p > .05). Group differences were found on English studying hours: intermediate learners had the longest L2 studying time; beginning learners had the shortest time; and advanced learners took the intermediate position, being significantly different from either group (all p < .05). These differences were derived from the natural groups’ design. Both age and sex were put in the LMMs for the overall RTs analysis, and no correlations were found (all p > .05) (see Table 1).

Main analyses

ANT

Overall performance (i.e., RTs) is illustrated in Figure 2. Fixed effects were not significant in the overall RTs (all p > .05) (marginal R² = .005, conditional R² = .464).

Alerting effect

No main effects or interactions were significant (all p > .05) (marginal R² = .028, conditional R² = .459).

Orienting effect

The orienting effect was significant, with faster responses on single-cue trials than on centre-cue trials (β = 45.43, 95% confidence interval [CI] = [5.80, 85.05], t = 2.25, p = .03]. No other fixed effects or interactions were significant (all p > .05) (marginal R² = .028, conditional R² = .459).

Conflict effect

The conflict effect was significant, with faster responses on congruent trials than on incongruent trials (β = 99.20, 95% CI = [76.83, 121.58], t = 8.89, p < .001]. No other fixed effects or interactions were significant (all p > .05) (marginal R² = .159, conditional R² = .487) (see Figure 3).

Switching effect

The switching effect was significant, with faster responses on non-switching trials than on switching trials (β = 8.56, 95% CI = [5.66, 11.47], t = 5.78, p < .001]. The interaction between the group and the switching effect was significant: The intermediate group obtained a smaller switching effect than the beginning group (β = 7.25, 95% CI = [−12.27, −2.21], t = 2.82, p = .005), with no other group differences (advanced vs. beginning, p = .33; advanced vs. intermediate, p = .11) (see Figure 4). No other fixed effects or interactions were significant (all p > 0.5) (marginal R² = .002, conditional R² = NA).

SCE

The main effects of the current trial and previous trial were significant, with faster responses on incongruent trials than on congruent trials (all p < .001). The two-way interaction between the previous trial and current trial was significant, with a larger congruency effect following congruent trials than following incongruent trials (i.e., SCE) (β = 25.80, 95% CI = [17.79, 33.80], t = 6.32, p < .001). The three-way interaction between the group and the SCE was significant: The intermediate group obtained a smaller SCE than the beginning group (β = 17.42, 95% CI = [3.54, 31.31], t = 2.50, p = .014), with no other group differences (advanced vs. beginning, p = .13; advanced vs. intermediate, p = .41) (see Figure 5). No other fixed effects or interactions were significant (all p > 0.5) (marginal R² = .159, conditional R² = .482).

Number Stroop task

Overall performance is illustrated in Figure 2. In the model for overall RTs, no fixed effects were significant (all p > .05) (marginal R² = .005, conditional R² = .418).

Stroop effect

The Stroop effect was significant, with faster responses on congruent trials than on incongruent trials (β = 75.78, 95% CI = [45.73, 105.84], t = 4.94, p < .001). The interaction between the group and the Stroop effect was significant: The advanced group obtained a smaller Stroop effect than the beginning group (β = 21.06, 95% CI = [6.73, 35.40], t = 2.88, p = .004), with no other group differences (intermediate vs. beginning, p = .65; advanced vs. intermediate, p = .036; adjusted p = .0167) (see Figure 3). No other fixed effects or interactions were significant (all p > 0.5) (marginal R² = .071, conditional R² = .433).

Switching effect

The switching effect was significant but reversed, with faster responses on switching trials than on non-switching trials (β = 5.24, 95% CI = [0.87, 9.61], t = 2.35, p = .014). No other fixed effects or interactions were significant (all p > 0.5) (marginal R² = .001, conditional R² = .415) (see Figure 4).

SCE

The main effects of the previous trial and current trial were significant, with faster responses on incongruent trials than on congruent trials (p < .001, p = .002, respectively). The two-way interaction between the previous trial and current trial was significant but reversed, with a larger congruency effect following incongruent trials than following congruent trials (i.e., SCE) (β = 17.13, 95% CI = [5.63, 28.63], t = 2.92, p = .004). The two-way interaction between the group and the current trial was significant: The advanced group obtained a smaller congruency effect than the beginning group (β = 25.16, 95% CI = [7.94, 42.38], t = 2.86, p = .004), with no other group differences (intermediate vs. beginning, p = .40; advanced vs. intermediate, p = .08). No other fixed effects or interactions were significant (all p > 0.5) (marginal R² = .117, conditional R² = .480) (see Figure 5).

TEA

There was no group difference in the ED (all p > 0.5) (R² = .003, R²_adjusted = −.014). A significant group effect was found in the ER: The intermediate group scored higher than the beginning group (Est = 15.53, 95% CI = [3.09, 27.97], t = 2.47, p = .015), with no other group differences (intermediate vs. advanced, p = .29; advanced vs. beginning, p = .21) (R² = .051, R²_adjusted = .035). Performance in the respective subtests is illustrated in Figure 6.

Further multiple stepwise analysis

Further multiple stepwise analysis suggested significant correlations between language-related variables, as well as demographic variables, and executive performance. The results (see the supporting materials) showed that language exposure, SES, and language proficiency were the three most significant predictors of performance across all cognitive measures, with significance level ranking from the strongest to the least. One exception was that nonverbal IQ, rather than language proficiency, ranked as the third significant predictor of performance in the ER.

Discussion

Given the foreign language policy in China, an overwhelming majority study English. This study set out to compare three groups of English learners in the Chinese population, who were relatively homogeneous in background variables (e.g., cultural and educational environment) but heterogeneous in their experience of L2 language acquisition (i.e., language proficiency, exposure, and usage). Three non-linguistic cognitive tasks were used to assess cognitive performance, tapping into multiple dimensions of executive functions, namely inhibitory control, switching ability, and attentional disengagement.

The results suggest that the bilingual experience benefitted the L2 learners in specific aspects of cognitive control in both the visual and auditory domains. In the visual domain, compared with beginning learners, advanced learners obtained greater inhibitory control in the Stroop task, and intermediate learners showed better-switching ability and faster engagement of attention in the ANT. In the auditory domain, intermediate learners showed superior performance in attentional switching relative to beginning learners in the ER subtest of the TEA. No group differences were found in other cognitive control as measured by the three tasks. Further multiple stepwise analyses suggested that language exposure and proficiency played significant modulating roles in cognitive control after controlling for background measures, such as SES, IQ, and working memory.

In the visual domain, better inhibitory control was found in the Stroop task by advanced learners than by beginning learners, which was consistent with previous studies (Costa et al., 2008; Hernández et al., 2010). Advanced and intermediate learners have similar self-reported language proficiency, but advanced learners had received much longer cumulative similar English learning/training experience as intermediate learners for at least 3 years. This longer cumulative language training experience indicates longer simultaneous exposure to two languages and a longer history of using two languages, along with higher L2 proficiency, which has been proposed to show greater cognitive benefits (Bialystok, 2017; Lehtonen et al., 2018). This could be the reason why the group difference in inhibitory performance was only observed between advanced learners and beginning learners.

Similarly, better-switching ability and attention disengagement were found in the ANT by intermediate learners than by beginning learners. At the testing time, intermediate learners received much more intensive L2 exposure than both advanced and beginning learners, as reflected in the weekly hours spent on English. As a result, the intermediate group might be more frequently switching between their two languages, thus leading to more efficient disengagement of attention and switching ability. This is confirmed by further multiple stepwise regression analyses that L2 exposure was the strongest predictor of cognitive performance. Previous studies have supported this hypothesis that a relatively intensive learning style predicts better language and cognitive performance (Serrano, 2011).

In the auditory domain, the findings are in line with previous research, which suggested that the cognitive effects associated with bilingualism were extended from the visual to the auditory domains (Bak et al., 2014; Vega-Mendoza et al., 2015). The absence of group differences in the ED could be because the ED was less demanding than the ER. Specifically, the ER task involves multiple cognitive processes, such as inhibiting low and high tones from counting while efficiently disengaging inhibition, monitoring the counting directions, and refocusing attention on the middle tone (Long et al., 2019).

Group differences in switching effects and SCEs were limited to the ANT, not the Stroop task, which could be due to the differences in response-to-stimulus intervals (RSIs) in the two tasks. Grundy, Chung-Fat-Yim, et al. (2017) observed behaviour differences between monolinguals and bilinguals in SCEs with their RSIs ranging from 250 to 1,000 ms, but the largest group difference occurred within 500 ms. The behavioural group difference disappeared when the RSIs were set from 1,000 to 1,500 ms, but the electrophysiological group difference occurred within 500 ms. They explained that bilinguals might have earlier disengagement of attention than monolinguals, within 500 ms. With longer RSI, both monolinguals and bilinguals had enough time to disengage from previous trials, leading to smaller group differences. In this study, the RSIs were 500 and 2,000 ms for the ANT and Stroop, respectively. Moreover, it is worthy of note that the switching effects and SCEs were reversed in the Stroop task in this study, this suggested that the RSI played a modulating role in performance on switching across trials and attentional disengagement from previous trials.

Group differences in conflict resolution were only observed in the Stroop task, not in the ANT. The possible reason might be that the group differences in the conflict effect were masked by the SCE (Grundy, Chung-Fat-Yim, et al., 2017). Standardised analyses of ANT networks are based on mean RTs to congruent and incongruent trials, without regard for the previous trials. Reliance on previous trial information could lead to two consequences in responses: faster responses when the previous and current trial types were the same (cC and iI) and slower responses when the trial types were different (cI and iC). Our findings revealed that intermediate learners had a greater ability to disengage attention from previous trials, which meant that they might experience less influence from the previous trials. This could average their performance and result in the reduction in the group difference concerning the conflict effect.

One potential limitation of this study is the gender distribution across the three groups. Given the unbalanced gender in the major selection, for example, there was a higher proportion of males in STEM majors (i.e., science, technology, engineering, and mathematics) and a higher proportion of females in non-STEM majors, especially in the arts and linguistics; it was, therefore, difficult to recruit an equal gender proportion in both groups. Secondary analyses were conducted on beginning learners to explore the potential effects of gender on these measured cognitive abilities. The results showed no differences between female and male participants in any of these cognitive measures, thus indicating that gender is not a significant predictor of cognitive performance in this study. Another potential limitation was the major of our participants; while most of the beginning learners were majoring in Physical Education, all the advanced and intermediate learners were majoring in English. This difference is consistent with the differences in the required overall university scores (e.g., English scores) between the two majors. Studies have suggested that regular physical exercise can ameliorate cognitive function in healthy young adults (see a review by Lubans et al., 2016). The differences in major and gender are due to the design of natural groups, mainly due to the limited access to ‘matched’ participants. A further potential limitation might be the lack of a formal test for English proficiency. However, studies have confirmed the validity of self-reported language proficiency and usage in bilingualism research (Luk & Bialystok, 2013; Marian et al., 2007). For instance, Vega-Mendoza et al. (2015) found a correlation between self-reported and objective assessments of language proficiency. Therefore, the self-reported language proficiency in this study is likely to be reliable (Xia et al., 2022).

Conclusion

The present findings suggest that early adult L2 learners experience similar bilingual effects reported in bilinguals across the lifespan, which are modulated by different language experience. Theoretically, this work suggests that future research into bilingual cognitive control should consider the characteristics of the selected tasks and focus on aspects of individuals’ specific language learning experience that might differently modulate language control and, in turn, affect cognitive functions. In addition, this study provides some bases for recommending to L2 young adult learners that language usage and language exposure are important in language learning.

Supplemental Material

sj-docx-1-ijb-10.1177_13670069241307606 – Supplemental material for The effect of language proficiency, usage, and exposure on cognitive control: A study in early adulthood Chinese learners of English

Supplemental material, sj-docx-1-ijb-10.1177_13670069241307606 for The effect of language proficiency, usage, and exposure on cognitive control: A study in early adulthood Chinese learners of English by Lihua Xia, Antonella Sorace, Mariana Vega-Mendoza, Xiaohong Deng and Thomas H. Bak in International Journal of Bilingualism

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Social Science Fund of China (grant no. 22CYY016); the Independent Innovation Fund of Huazhong University of Science and Technology (grant no. 2022WKYXQN005); and support from the School of Philosophy, Psychology & Language Sciences (PPLS) Research Support Grants.

ORCID iDs

Lihua Xia

Mariana Vega-Mendoza

Supplemental material

Supplemental material for this article is available online.

Notes

Author biographies

Lihua Xia is a researcher specializing in bilingualism, with a particular focus on how bilingualism affects cognitive processes across the lifespan. Her work examines the cognitive and neural mechanisms that underlie bilingualism, exploring how different stages of language acquisition and bilingual experience influence cognitive development and decline over time.

Antonella Sorace, FBA, FRSE, FRSA, is an experimental linguist and academic, specializing in bilingualism across the lifespan. Her research has contributed significantly to understanding the cognitive and neural processes involved in managing multiple languages.

Mariana Vega-Mendoza is a researcher specializing in psycholinguistics, bilingualism, and cognitive science. She has contributed to understanding the cognitive mechanisms underlying bilingualism and how language learning impacts cognition.

Xiaohong Deng is a professor at Hubei University, specializing in cognitive neuroscience with a focus on the application of advanced technical tools such as EEG in psychological research. Her research interests include cognitive processing, neural mechanisms of behaviour, and the use of neuroimaging techniques to explore brain functions.

Thomas H. Bak is a well-known researcher in the fields of cognitive science, linguistics, and psychology, particularly known for his work on bilingualism and cognitive aging. He has explored the impact of bilingualism on cognitive reserve and brain health, often investigating how speaking more than one language might delay the onset of Alzheimer’s disease and other cognitive decline conditions.

References

Bak

T. H.

Long

M. R.

Vega-Mendoza

Sorace

(2016). Novelty, challenge, and practice: The Impact of intensive language learning on attentional functions. PLOS ONE, 11(4), Article e0153485. https://doi.org/10.1371/journal.pone.0153485

Bak

T. H.

Vega-Mendoza

Sorace

(2014). Never too late? An advantage on tests of auditory attention extends to late bilinguals. Frontiers in Psychology, 5, Article 485. https://doi.org/10.3389/fpsyg.2014.00485

Barton

(2014). MuMIn: Multi-model inference (R Package Version 1.10.5). http://CRAN. R-project. org/package= MuMIn

Bates

Mächler

Bolker

Walker

(2015). Fitting linear mixed-effects models using lme4. Journal of Statistical Software, 67(1), 1–48. https://doi.org/10.18637/jss.v067.i01

Bialystok

(2017). The bilingual adaptation: How minds accommodate experience. Psychological Bulletin, 143(3), 233–262. https://doi.org/10.1037/bul0000099

Bialystok

(2020). Null results in bilingualism research: What they tell us and what they don’t. Journal of Multilingual Theories and Practices, 1(1), 8–22.

Bialystok

Craik

F. I. M.

Luk

(2012). Bilingualism: Consequences for mind and brain. Trends in Cognitive Sciences, 16(4), 240–249. https://doi.org/10.1016/j.tics.2012.03.001

Blumenfeld

H. K.

Marian

(2011). Bilingualism influences inhibitory control in auditory comprehension. Cognition, 118(2), 245–257. https://doi.org/10.1016/j.cognition.2010.10.012

Bonfieni

Branigan

H. P.

Pickering

M. J.

Sorace

(2019). Language experience modulates bilingual language control: The effect of proficiency, age of acquisition, and exposure on language switching. Acta Psychologica, 193, 160–170. https://doi.org/10.1016/j.actpsy.2018.11.004

10.

Botvinick

M. M.

Braver

T. S.

Barch

D. M.

Carter

C. S.

Cohen

J. D.

(2001). Conflict monitoring and cognitive control. Psychological Review, 108(3), 624–652. https://doi.org/10.1037/0033-295X.108.3.624

11.

Costa

Hernández

Sebastián-Gallés

(2008). Bilingualism aids conflict resolution: Evidence from the ANT task. Cognition, 106(1), 59–86. https://doi.org/10.1016/j.cognition.2006.12.013

12.

Degirmenci

M. G.

Grossmann

J. A.

Meyer

Teichmann

(2022). The role of bilingualism in executive functions in healthy older adults: A systematic review. International Journal of Bilingualism, 26(4), 426–449. https://doi.org/10.1177/13670069211051291

13.

Egner

(2007). Congruency sequence effects and cognitive control. Cognitive, Affective, & Behavioral Neuroscience, 7(4), 380–390. https://doi.org/10.3758/CABN.7.4.380

14.

Fan

McCandliss

B. D.

Sommer

Raz

Posner

M. I.

(2002). Testing the Efficiency and Independence of Attentional Networks. Journal of Cognitive Neuroscience, 14(3), 340–347. https://doi.org/10.1162/089892902317361886

15.

Gratton

Coles

M. G. H.

Donchin

(1992). Optimizing the use of information: Strategic control of activation of responses. Journal of Experimental Psychology: General, 121(4), 480–506. https://doi.org/10.1037/0096-3445.121.4.480

16.

Green

D. W.

Abutalebi

(2013). Language control in bilinguals: The adaptive control hypothesis. Journal of Cognitive Psychology, 25(5), 515–530. https://doi.org/10.1080/20445911.2013.796377

17.

Grosjean

(2010). Bilingualism, biculturalism, and deafness. International Journal of Bilingual Education and Bilingualism, 13(2), 133–145. https://doi.org/10.1080/13670050903474051

18.

Grundy

J. G.

(2020). The effects of bilingualism on executive functions: An updated quantitative analysis. Journal of Cultural Cognitive Science, 4(2), 177–199. https://doi.org/10.1007/s41809-020-00062-5

19.

Grundy

J. G.

Anderson

J. A. E.

Bialystok

(2017). Neural correlates of cognitive processing in monolinguals and bilinguals. Annals of the New York Academy of Sciences, 1396, 183–201. https://doi.org/10.1111/nyas.13333

20.

Grundy

J. G.

Chung-Fat-Yim

Friesen

D. C.

Mak

Bialystok

(2017). Sequential congruency effects reveal differences in disengagement of attention for monolingual and bilingual young adults. Cognition, 163, 42–55. https://doi.org/10.1016/j.cognition.2017.02.010

21.

Grundy

J. G.

Keyvani Chahi

(2017). Post-conflict slowing effects in monolingual and bilingual children. Developmental Science, 20(1), Article e12488. https://doi.org/10.1111/desc.12488

22.

Gullifer

J. W.

Chai

X. J.

Whitford

Pivneva

Baum

Klein

Titone

(2018). Bilingual experience and resting-state brain connectivity: Impacts of L2 age of acquisition and social diversity of language use on control networks. Neuropsychologia, 117, 123–134. https://doi.org/10.1016/j.neuropsychologia.2018.04.037

23.

Hernández

Costa

Fuentes

L. J.

Vivas

A. B.

Sebastián-Gallés

(2010). The impact of bilingualism on the executive control and orienting networks of attention. Bilingualism: Language and Cognition, 13(3), 315–325. https://doi.org/10.1017/S1366728909990010

24.

Hui

N. -Y.

Yuan

Fong

M. C. -M.

Wang

W. S.

(2020). L2 proficiency predicts inhibitory ability in L1-dominant speakers. International Journal of Bilingualism, 24(5–6), 984–998. https://doi.org/10.1177/1367006920914399

25.

Kapa

L. L.

Colombo

(2013). Attentional control in early and later bilingual children. Cognitive Development, 28(3), 233–246. https://doi.org/10.1016/j.cogdev.2013.01.011

26.

Kroll

Dussias

Bice

Perrotti

(2015). Bilingualism, mind, and brain. Annual Review of Linguistics, 1(1), 377–394. https://doi.org/10.1146/annurev-linguist-030514-124937

27.

Kroll

J. F.

Bialystok

(2013). Understanding the consequences of bilingualism for language processing and cognition. Journal of Cognitive Psychology, 25(5), 37–41. https://doi.org/10.1080/20445911.2013.799170

28.

Lehtonen

Soveri

Laine

Järvenpää

de Bruin

Antfolk

(2018). Is bilingualism associated with enhanced executive functioning in adults? A meta-analytic review. Psychological Bulletin, 144(4), 394–425. https://doi.org/10.1037/bul0000142

29.

Long

M. R.

Vega-Mendoza

Rohde

Sorace

Bak

T. H.

(2019). Understudied factors contributing to variability in cognitive performance related to language learning. Bilingualism: Language and Cognition, 23, 801–811. https://doi.org/10.1017/S1366728919000749

30.

Lubans

Richards

Hillman

Faulkner

Beauchamp

Nilsson

Kelly

Smith

Raine

Biddle

(2016). Physical activity for cognitive and mental health in youth: A systematic review of mechanisms. Pediatrics, 138(3), Article e20161642. https://doi.org/10.1542/peds.2016-1642

31.

Luk

Bialystok

(2013). Bilingualism is not a categorical variable: Interaction between language proficiency and usage. Journal of Cognitive Psychology, 25(5), 605–621. https://doi.org/10.1080/20445911.2013.795574

32.

Luk

De Sa

Bialystok

(2011). Is there a relation between onset age of bilingualism and enhancement of cognitive control? Bilingualism: Language and Cognition, 14(4), 596–596. https://doi.org/10.1017/S1366728911000186

33.

Marian

Blumenfeld

H. K.

Kaushanskaya

(2007). The Language Experience and Proficiency Questionnaire (LEAP-Q): Assessing language profiles in bilinguals and multilinguals. Journal of Speech, Language, and Hearing Research, 50, 940–967. https://doi.org/10.1044/1092-4388(2007/067)

34.

Mishra

R. K.

(2015). Let’s not forget about language proficiency and cultural variations while linking bilingualism to executive control. Bilingualism: Language and Cognition, 18(1), 39–40. https://doi.org/10.1017/S1366728914000455

35.

Mishra

R. K.

Hilchey

M. D.

Singh

Klein

R. M.

(2012). On the time course of exogenous cueing effects in bilinguals: Higher proficiency in a second language is associated with more rapid endogenous disengagement. Quarterly Journal of Experimental Psychology, 65(8), 1502–1510. https://doi.org/10.1080/17470218.2012.657656

36.

Paap

K. R.

Anders-Jefferson

Zimiga

Mason

Mikulinsky

(2020). Interference scores have inadequate concurrent and convergent validity: Should we stop using the flanker, Simon, and spatial Stroop tasks? Cognitive Research: Principles and Implications, 5(1), 7. https://doi.org/10.1186/s41235-020-0207-y

37.

Paap

K. R.

Greenberg

Z. I.

(2013). There is no coherent evidence for a bilingual advantage in executive processing. Cognitive Psychology, 66(2), 232–258. https://doi.org/10.1016/j.cogpsych.2012.12.002

38.

Raven

J. C.

Foulds

G. A.

(1962). Advanced progressive matrices. H.K. Lewis & Co. Ltd.

39.

Robertson

I. H.

Ward

Ridgeway

Nimmo-Smith

(1994). The Test of Everyday Attention (TEA). Thames Valley Test Company.

40.

Serrano

(2011). The time factor in EFL classroom practice. Language Learning, 61(1), 117–145. https://doi.org/10.1111/j.1467-9922.2010.00591.x

41.

Sulpizio

Del Maschio

Fedeli

Abutalebi

(2020). Bilingual language processing: A meta-analysis of functional neuroimaging studies. Neuroscience and Biobehavioral Reviews, 108, 834–853. https://doi.org/10.1016/j.neubiorev.2019.12.014

42.

Thierry

Y. J.

(2007). Brain potentials reveal unconscious translation during foreign-language comprehension. Proceedings of the National Academy of Sciences of the United States of America, 104(30), 12530–12535. https://doi.org/10.1073/pnas.0609927104

43.

Treccani

Argyri

Sorace

Della Sala

(2009). Spatial negative priming in bilingualism. Psychonomic Bulletin & Review, 16(2), 320–327. https://doi.org/10.3758/PBR.16.2.320

44.

Unsworth

(2015). Amount of exposure as a proxy for dominance in bilingual language acquisition. In Silva-Corvalán

Treffers-Daller

(Eds.), Language dominance in bilinguals: Issues of measurement and operationalization (pp. 156–173). Cambridge University Press.

45.

Vega-Mendoza

West

Sorace

Bak

T. H.

(2015). The impact of late, non-balanced bilingualism on cognitive performance. Cognition, 137, 40–46. https://doi.org/10.1016/j.cognition.2014.12.008

46.

Ware

A. T.

Kirkovski

Lum

J. A. G.

(2020). Meta-analysis reveals a bilingual advantage that is dependent on task and age. Frontiers in Psychology, 11, Article 1458. https://doi.org/10.3389/fpsyg.2020.01458

47.

Wechsler

(1997). Wechsler Memory Scale (WMS-III) (Vol. 14). Psychological Corporation.

48.

Xia

Bak

Sorace

Vega-Mendoza

(2022). Interference suppression in bilingualism: Stimulus-stimulus vs. stimulus-response conflict. Bilingualism: Language and Cognition, 25(2), 256–268. https://doi.org/10.1017/S1366728921000304

49.

Yurtsever

Anderson

J. A. E.

Grundy

J. G.

(2023). Bilingual children outperform monolingual children on executive function tasks far more often than chance: An updated quantitative analysis. Developmental Review, 69, 101084. https://doi.org/10.1016/j.dr.2023.101084

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.06 MB