![Sample our Language & Literature journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5941/TOC-Subject-Sample-2021-Language-Literature.jpg"})
Abstract
This exploratory study investigated executive attention during nonverbal and verbal processing among adults with a range of bilingual experience. Previous research has found that bilingual children control their attention better than their monolingual peers and that superior attentional control in some processing contexts persists into adulthood among lifelong bilinguals. An open question is whether late-acquired experience learning and using two languages can lead to enhanced executive attention or whether these cognitive advantages are available only to individuals whose bilingualism develops in early childhood. A total of 48 Spanish–English bilinguals completed verbal and nonverbal tasks designed to assess aspects of executive attention including inhibitory control and monitoring and switching of attention (i.e. working memory capacity). Preliminary results suggest an association between bilingual experience and enhanced efficiency of these components of executive attention in the nonverbal domain. Furthermore, a significant relationship between the efficiency of inhibitory control in verbal and nonverbal domains hints at a connection between specific control over language and enhanced domain-general executive control that may be more or less evident depending on task demands. These encouraging pilot results warrant larger-scale replication that brings together data from both linguistic and nonlinguistic processing as bilingualism research seeks to extend its understanding of bilingual cognition.
Acknowledgements
Many thanks to two anonymous reviewers for their thoughtful and helpful comments on an earlier version of this paper; any remaining errors are my own. I also most gratefully acknowledge Cristina Sanz of Georgetown University for access to the online working memory tests, which were computer-adapted under the auspices of The Latin Project, as well as Bill Garr and Gorky Cruz, also of Georgetown University, for technical support. The research reported in this paper was supported by a grant from the University of Wisconsin-Madison Graduate School with funding from the Wisconsin Alumni Research Foundation.
Correspondence concerning this article should be addressed to the author, Department of Spanish and Portuguese, University of Wisconsin-Madison, 1018 Van Hise Hall, 1220 Linden Drive, Madison, WI 53706, USA.
Notes
1. ‘Simple’ is used here to contrast these tasks with dual-task measures of WM capacity such as those employed in the present study, in which participants’ attention is divided as they perform concurrent tasks, thus enabling evaluation of the monitoring and switching functions of executive attention.
2. Acknowledging that for these six participants English would be more appropriately characterized as a second L1, for the sake of consistency Spanish will be referred to as the L1 and English as the L2 throughout the remainder of the paper.
3. Points on the proficiency scales were as follows: 1=none or almost none; 2=poor; 3=functional; 4=good; 5=very good; and 6=like a native speaker.
4. As one reviewer pointed out, the small number of trials is a limitation as between-groups differences in RTs are less apt to be reliable. While the idea of completing just one trial block was to prevent practice effects such as those that have been experienced in other studies, we certainly acknowledge the limitation and in a large-scale study would seek more balance between ensuring a more robust data-set and keeping participants’ performance unpracticed enough so as to avoid the development of test-taking strategies.
5. Consistent with span tests that have been used in cognitive psychology research, sets range in length from three to five sentences so that participants cannot anticipate when they will be cued to complete the recall task.
6. In their analysis of the validity and reliability of verbal WM tests that differed in their inclusion of one (a sense judgment) or two (a sense and a grammaticality judgment) processing components, Winke, Stafford, and Adams (Citation2003) found that the measures that included two judgments had superior discriminatory power. It was therefore decided to include two processing components in the WM tests employed in this study, and the English listening span test used in this study is an adaptation of the test used in Mackey et al.'s (2010) study and is the more highly discriminatory measure of those analyzed by Winke, Stafford, and Adams (Citation2003). The Spanish test was developed by the researcher.
7. While our original intention was to have 12 of each sentence type on both span tests, misidentification of types or inadvertent misreading during the recording of the tests led to the following distributions by sentence type: (on the L1 Spanish test) 13 +S+G, 10 +S−G, 12 −S+G, 13 −S−G; (on the L2 English test) 12 +S+G, 12 +S−G, 13 −S+G, 11 −S−G.
8. The Pearson correlation coefficients were as follows: L1 plausibility×recall, r(46)=−0.03; L2 plausibility×recall, r(45)=−0.04; L2 grammar×recall, r(45)=0.12, all ps>0.05. Interestingly, there was a significant, positive correlation between L1 grammar judgment accuracy and L1 recall, r(46)=0.35, p=0.02.
9. It is acknowledged that some researchers (e.g. Conway et al. Citation2005; Cowan et al. Citation2005) argue that it is sufficient to include a cognitively demanding secondary task in the design of a dual-task span test to ensure that attention is being divided and that performance on the secondary task need not be taken into consideration in scoring, while other researchers (e.g. Turner and Engle Citation1989; Mackey et al. Citation2010; Walter Citation2004) have suggested that a more adequate measure of WM capacity in a dual-task paradigm is the sum of participants’ performance on both tasks (i.e. processing and storage). It is beyond the scope and not the intention of this study to provide a detailed evaluation of differing operationalizations of WM capacity (see Conway et al. Citation2005 for thoughtful discussion of this issue); however, as an exploratory aside, scores were calculated according to all three methods and scores were significantly and strongly correlated (rs>0.90), suggesting a high degree of overlap in what the various operationalizations tap into.
10. The data for the study reported here were collected as part of a larger project which is detailed elsewhere. Participants completed other tasks at each of the study sessions, but data which are not central to the current study's research questions are not discussed here.
11. The author thanks an anonymous reviewer for pointing out this parallel.
12. Although seven adults over 40 were included in the present sample, the average AoT was 32 years, and in order to hold current age constant in the LEB and MEB groups, all participants included in the ANOVA groups were in their 20s and 30s.