Elsevier

Cognition

Volume 113, Issue 2, November 2009, Pages 135-149
Cognition

On the bilingual advantage in conflict processing: Now you see it, now you don’t

https://doi.org/10.1016/j.cognition.2009.08.001Get rights and content

Abstract

We report two experiments exploring more in detail the bilingual advantage in conflict resolution tasks. In particular, we focus on the origin of the bilingual advantage on overall reaction times in the flanker task. Bilingual and monolingual participants were asked to perform a flanker task under different task versions. In Experiment 1, we used two low-monitoring versions where most of the trials were of just one type (either congruent or incongruent). In Experiment 2, we used two high-monitoring versions where congruent and incongruent trials were more evenly distributed. An effect of bilingualism in overall reaction times was only present in the high-monitoring condition. These results reveal that when the task at hand recruits a good deal of monitoring resources, bilinguals outperform monolinguals. This observation suggests that bilingualism may affect the monitoring processes involved in executive control.

Introduction

Bilingualism has been shown to have a positive effect on the functioning of an individual’s attentional system across the life-span (e.g., Bialystok, 1999, Bialystok and Martin, 2004, Carlson and Meltzoff, 2008, Costa et al., 2008, Martin-Rhee and Bialystok, 2008; see full list of studies in Appendix A). The bilingual advantage has been explored mainly by means of tasks that involve ignoring distracting and conflicting information, such as the Simon task and the flanker task. There are two effects that have been taken as indexing a bilingual advantage:

  • (a)

    The conflict produced by incongruent distractors is larger for monolinguals than for bilinguals (e.g., Bialystok et al., 2004, Bialystok et al., 2008, Costa et al., 2008).

  • (b)

    Bilinguals perform tasks involving conflict resolution overall faster than monolinguals (e.g., Bialystok, 2006, Costa et al., 2008, Martin-Rhee and Bialystok, 2008).

Consider, for example, the flanker task in which participants are asked to indicate whether a central arrow (→) points to the right or to the left. This arrow is presented along with four flanker arrows pointing to the same (congruent trials →→→→→) or different direction (incongruent trials ←←→←←). Responses tend to be slower for incongruent than for congruent trials (the conflict effect), revealing the time needed to resolve the conflict between the target and the flankers. The bilingual advantage in this task was indexed by a reduced conflict effect for bilinguals in comparison to monolinguals (Costa et al., 2008). Interestingly, bilinguals were also overall faster in performing the task. That is, bilinguals in comparison to monolinguals responded faster both in congruent and incongruent trials. This effect may reflect an impact of bilingualism on the efficiency of other cognitive processes than conflict resolution per se, as has already been suggested by several authors (e.g., response suppression and switching; Bialystok, Craik, & Ryan, 2006).

The reliability and robustness of the two effects associated to bilingualism mentioned above appear to be somewhat different. A review of the literature is complex, given the different types of populations, tasks and designs used. However, in a large number of studies the magnitude of the conflict effect is similar for bilinguals and monolinguals (see Appendix A). Most of these studies have used mixed designs (25/37 studies) in which congruent and incongruent trials were randomly presented. Out of these 25 experiments only 6 led to a bilingual advantage in the magnitude of the conflict effect. However, there was a significant difference in overall reaction times (RTs) between the two groups in about 12 of these 25 experiments. It is worth mentioning that most of these studies used mixing contexts in which congruent and incongruent trials were presented evenly (22/25), and in only 12 of the 22 there was a bilingual effect whatsoever. Importantly, however, when the design is blocked differences associated to bilingualism are extremely difficult to detect (3/12).

Two empirical generalizations can be derived from these results. A bilingual advantage is more likely to be present in: (a) overall RTs than on the magnitude of the conflict effect, (b) in mixing contexts where congruent and incongruent trials are randomly presented, than in blocked designs. In this article, we explore the relationship between these two empirical generalizations. We do so by assessing how the percentage of congruent and incongruent trials in mixed designs affect the detectability of the bilingual impact on both overall RTs and the magnitude of conflict effects.

Section snippets

On the origin of the Bilingual advantage on overall RTs and on the conflict effect

From the theoretical point of view, the bilingual advantage on the magnitude of the conflict effect and on the overall RTs may index the impact of bilingualism on different components of the attentional system. The first effect may be related to processes recruited in suppressing interference from irrelevant conflicting information. The need of bilinguals to continuously control their two languages, focusing on the relevant linguistic representation while avoiding interference from the

The present study

In this study we test the hypothesis that the bilingual advantage in overall RTs stems from a more efficient monitoring system. The rationale of the study is the following: if the bilingual advantage is in some way related to the functioning of the monitoring system, then it should be present in conditions requiring high monitoring demands, and reduced or absent in those condition in which the monitoring system is less taxed. To manipulate the involvement of the monitoring system, the

General method

In this version of the ANT

Experiment 1: Is there a bilingual advantage in low monitoring contexts?

This experiment included two tasks versions. In the first version, most of the trials were congruent (92% congruent version −88 congruent trials and 8 incongruent trials per block) while in the second version most of them were incongruent (8% congruent version −8 congruent trials and 88 incongruent trials per block). An effect of bilingualism will be indexed either by a main effect of “Group of Participants” or by an interaction of this variable with the “Type of flanker” variable, since this

Results and discussion

RTs faster than 200 ms or slower than 1200 ms were excluded from the analyses (0.52% of the trials). We conducted a general analysis of variance (ANOVA) with “Block” (1 vs. 2 vs. 3) and “Flanker type” (congruent vs. incongruent) as within-subject factors, and “Group of Participants” (bilingual vs. monolingual) and “Task version” (8% congruent version vs. 92% congruent version) as between-subjects factors (see Appendix C for a description of the orienting and alerting effects).

In the RTs analyses

Experiment 2: Is there a bilingual advantage in high monitoring contexts?

This experiment included two tasks versions. In the first version, half of the trials were congruent and the other half incongruent (50% congruent version −48 congruent trials and 48 incongruent trials per block). In the second version, only 25% of the trials were incongruent (75% congruent version −72 congruent trials and 24 incongruent trials per block). An effect of bilingualism would be indexed either by a main effect of “Group of participants” in any of these versions or by an interaction

Results

The same criteria as in Experiment 1 to remove outliers were applied here (0.4% of the trials). Also, the same general analysis of variance (ANOVA) as in the previous study was performed.

In the RT analyses the main effect of “Task version” was not significant. The main effects of “Block” (F(2, 240) = 4.63, MSE = 921.93, p = .011) and “Flanker type” (F(1, 120) = 1482.84, MSE = 1667.65, p < .0001) were significant. Crucially, the main effect of “Group of Participants” was significant (F(1, 120) = 11.86, MSE = 

Discussion

Several interesting results were observed in this experiment. First, bilinguals tended to be overall faster than monolinguals (Fig. 2, Panel B). This effect was clearly significant across the whole experiment in the condition in which the monitoring system is maximally recruited (50% congruent version) and was present

General discussion

The experiments reported in this article aimed at exploring some of the conditions that lead to a bilingual advantage in tasks that involve conflict processing. In particular, we focused on the origin of the bilingual advantage in overall speed. We hypothesized that such an effect might reveal an impact of bilingualism on the efficiency of the conflict monitoring system, and that consequently, its detectability would depend on the extent to which the experimental conditions tax the monitoring

The bilingual advantage on overall RTs

One of the indexes of a bilingual advantage in conflict resolution tasks is the overall faster RTs of bilinguals in comparison to monolinguals. Arguably, this effect cannot be attributed merely to a more efficient conflict resolution processing, given that no conflict resolution is required when dealing with congruent stimuli (target and flankers call for the same response). We hypothesized that this bilingual advantage could be the result of a more efficient monitoring processing system, in

The bilingual advantage on the magnitude of the conflict effect

The results presented here are also relevant for the other signature effect of a bilingual advantage in conflict processing tasks, namely the reduction in the magnitude of the conflict effect in bilinguals. Although this effect has received much attention, it appears not to be very robust (see Appendix A), leading some authors to question its usefulness for drawing strong conclusions (Colzato et al., 2008). Our results add evidence to this concern. This is because a difference in the magnitude

Language processing, conflict resolution and conflict monitoring

In the Introduction we argued that the advantage in the magnitude of the conflict effect may stem from the need of bilinguals to lexicalize words in the intended language while preventing massive interference from the other language. There are several proposals regarding how this lexicalization is achieved, but all of them agree on assuming that some sort of language control is needed (Costa et al., 2006, Finkbeiner et al., 2006, Finkbeiner et al., 2006, Green, 1998, La Heij, 2005) (in fact, it

Acknowledgements

This research was supported by two grants from the Spanish Government (PSI2008-01191/PSIC with EC Fondos FEDER Consolider Ingenio 2010 CE-CSD2007-00121). Mireia Hernández was supported by a Pre-doctoral fellowship from the Catalan Government. (SGR-2005). We thank Elin Runnquist for her comments.

References (31)

  • E. Bialystok et al.

    Attention and inhibition in bilingual children: Evidence from the dimensional change card sort task

    Developmental Science

    (2004)
  • E. Bialystok et al.

    Bilingualism across the lifespan: The rise and fall of inhibitory control

    International Journal of Bilingualism

    (2005)
  • S.M. Carlson et al.

    Bilingual experience and executive 651 functioning in young children

    Developmental Science

    (2008)
  • L.S. Colzato et al.

    How does bilingualism improve executive control? A comparison of active and reactive inhibition mechanisms

    Journal of Experimental Psychology: Learning, Memory and Cognition

    (2008)
  • A. Costa et al.

    The dynamics of bilingual lexical access

    Bilingualism: Language and Cognition

    (2006)
  • Cited by (585)

    View all citing articles on Scopus
    View full text