The present study sought to replicate prior findings on a bilingual advantage in cognitive flexibility as well as investigate the association of bilingualism with attentional fluctuations. Previous studies have not considered consistency in attentional control when exploring the bilingual advantage. Our results showed no evidence for a bilingual advantage in either cognitive flexibility or attentional fluctuations, with bilingualism investigated both continuously and categorically. Furthermore, the nonsignificant results obtained with classical frequentist statistical methods were confirmed to stem from true null effects of bilingual experience on attentional fluctuations and cognitive flexibility by means of Bayes factor analyses.
Correlational analyses did not show any significant associations between continuous variables indexing bilingualism (L2 exposure and age of acquisition) and both outcome variables (cognitive flexibility and attentional fluctuations). Amount of L2 exposure also did not correlate with SES—this was likely because of the restricted range of our sample consisting of only middle- to high-SES families. Regression analyses also showed that L2 exposure did not significantly predict attentional fluctuations or cognitive flexibility, controlling for age of acquisition, race, and SES. Using Bayes factor analyses, we showed that L2 exposure predicted the data substantially worse [54
] than an intercept-only model, adding to the evidence of no relationship between bilingualism and attentional fluctuations or cognitive flexibility. We included race and SES as covariates, given that they often confounded reported bilingual advantages [55
]. In the present study, race and SES were not significantly associated with cognitive flexibility or attentional fluctuations—however, race was associated with L2 exposure and AoA. We interpret the lack of associations with race, SES, and study outcomes to be due to our specific sample, and we still encourage future studies to treat these variables as confounds.
The hypothetical link between bilingualism and cognitive flexibility derives from the notion that bilinguals must constantly attend to one language and flexibly shift to the other language when prompted. This practice in shifting the relative activation of language systems in working memory is believed to lead to improved performance on cognitive flexibility tasks [57
]. However, the present study found no evidence for a bilingual advantage in cognitive flexibility, concurring with previous null findings on this association [4
]. Our results may not match those studies purporting a bilingual advantage because of sample size—Paap et al. [7
] found that a bilingual advantage in cognitive flexibility was more likely in studies with smaller samples (N
< 30). However, it is notable that our categorical analysis included a similarly small sample size and still found no evidence for a bilingual advantage. Our null results could also be attributable to measuring bilingualism continuously—indeed, previous studies measuring degree of bilingualism failed to replicate findings from studies that indexed bilingualism dichotomously [29
]. To address this, we split our sample into a monolingual and bilingual group, with the bilingual group defined as those with continuous second language exposure since birth. This analysis still failed to find a group difference on cognitive flexibility. There may be other influential sample characteristics if a bilingual advantage is more likely to occur, such as age. Previous studies have suggested that a bilingual advantage on cognitive flexibility is more likely to present in children and the elderly, as opposed to young adults [63
]; however, we still did not find this result in our sample of children, which is in line with other studies showing null results on a bilingual advantage in children [55
Our study also did not find any associations between bilingualism and attentional fluctuations. Our interest in investigating this association arose out of the theory that the bilingual experience enhances overall attention processes (including selective and sustained attention) in addition to cognitive flexibility [22
]. According to this theory, bilinguals must constantly attend to two language systems and direct attention to task-relevant information to select which language to deploy [22
]. If this theory is true, one would expect the enhancement of attention processes to include reduced lapses in sustained attention (i.e., lower attentional fluctuations). Furthermore, given that attentional fluctuations are shown to underlie accuracy on cognitive tasks [12
], it may be that attentional fluctuations could modulate or mediate the relationship between bilingualism and cognitive flexibility. Indeed, attentional fluctuations and cognitive flexibility were significantly correlated in our sample, with more fluctuations in attention on one task associated with lower performance on the cognitive flexibility task. However, given that bilingualism did not relate to either of these outcome variables, we did not investigate any mediating or moderating relationships.
The present findings—as well as previous studies on this topic—are limited in interpretation because there is a lack of straightforward accounts on mechanisms underlying the supposed bilingual advantage. There is a general account that management of two languages enhances cognitive control and attention in bilinguals, but there is a lack of specific and falsifiable hypotheses in this area [66
]. Indeed, one researcher has likened studies on the bilingual advantage to “a hunt for treasure by randomly digging holes in uninhabited islands” [67
] (p. 337). An analogous debate is that on mechanisms underlying transfer in cognitive training programs [68
], it is unclear how transfer occurs, limiting research findings to post hoc explanations. We acknowledge that the present study also operates on a general theory connecting bilingualism to cognitive flexibility and attention. We do not investigate any specific mechanistic account, except with intention to explore whether attentional fluctuations might mediate or moderate between bilingualism and cognitive flexibility. Future studies—and the field as a whole—would benefit from testing specific accounts of language control and cognitive transfer [67
Several other limitations of the present study should be acknowledged. First, we used summary scores (amount of L2 exposure, age of acquisition) for our measurement of bilingualism. This did not allow for more fine-grained analysis of the frequency of language switching in individual children, which may be important given theory linking this variable to cognitive advantages. One promising new way to index bilingual language switching behavior involves collecting ecological momentary assessment data on this variable several times a day with smartphones, a methodology that has been recently developed [69
Another limitation pertaining to measurement was that we were unable to obtain standardized proficiency measures for the range of second languages in our sample. In our categorical analysis, we classified bilinguals as those with second language exposure since birth; however, we acknowledge that this may not necessarily map onto language proficiency. In order to address this issue, we performed an additional set of analyses (see Supplementary Materials
) where only Spanish L2 participants were considered for whom L1 and L2 proficiency scores (Peabody Picture Vocabulary Test, PPVT [70
]; Test de Vocabulario en Imagenes Peabody, TVIP [71
]) were available (N
= 37). We found no significant correlations between children’s L2 proficiency, attentional fluctuations, or cognitive flexibility. How balanced participants were in English and Spanish (determined by means of language dominance index, see Supplementary Materials
for further information) also did not correlate with their TEC or DCCS scores.
A second limitation is that our sample may restrict the generalizability of our findings. Our participants consisted of children who were all native English speakers, were born in the U.S., resided in Northern California, and were moderate-to-high SES. Results might be different for children of low SES, immigrant status, or in different regions of the country or world.
In summary, this study adds to null findings on the bilingual advantage in cognitive flexibility. Our results also found no evidence for an association between bilingualism and attentional fluctuations—an important contributor to cognitive performance. Because of this, we did not investigate whether attentional fluctuations moderated or mediated between bilingualism and cognitive flexibility. We found the same results when investigating bilingualism continuously and categorically. We interpret our results to mean that a bilingual advantage is not evident for individuals with characteristics and circumstances similar to our sample. Despite the present null results, we want to emphasize that ours, along with any other results reporting null effects of the bilingual experience on general cognition, should in no way discourage the development of dual-language proficiency and second language learning. Knowing another language brings countless advantages outside of the realms of cognition, which include (but are not limited to) one own’s cultural expansion, broadening of the horizons, open-mindedness and expanded communicative abilities. Future studies should test more specific mechanistic accounts of the interplay between language and cognition, as well as the value of language outside of cognitive domains, which will be informative for understanding the world’s growing population of bilingual individuals.