Musical Expertise, Bilingualism, and Executive Functioning

Transcription

1 Journal of Experimental Psychology: Human Perception and Performance 009, Vol. 35, No., American Psychological Association /09/$1.00 DOI: /a Musical Expertise, Bilingualism, and Executive Functioning Ellen Bialystok York University and Rotman Research Institute of Baycrest Anne-Marie DePape York University and McMaster University The authors investigated whether intensive musical experience leads to enhancements in executive processing, as has been shown for bilingualism. Young adults who were bilinguals, musical performers (instrumentalists or vocalists), or neither completed 3 cognitive measures and executive function tasks based on conflict. Both executive function tasks included control conditions that assessed performance in the absence of conflict. All participants performed equivalently for the cognitive measures and the control conditions of the executive function tasks, but performance diverged in the conflict conditions. In a version of the Simon task involving spatial conflict between a target cue and its position, bilinguals and musicians outperformed monolinguals, replicating earlier research with bilinguals. In a version of the Stroop task involving auditory and linguistic conflict between a word and its pitch, the musicians performed better than the other participants. Instrumentalists and vocalists did not differ on any measure. Results demonstrate that extended musical experience enhances executive control on a nonverbal spatial task, as previously shown for bilingualism, but also enhances control in a more specialized auditory task, although the effect of bilingualism did not extend to that domain. Keywords: music expertise, bilingualism, plasticity, executive control The inevitable trajectory of the brain as it loses volume through the reduction of both gray and white matter begins surprisingly early in life and is well documented (review in Buckner, Head, & Lustig, 006). Against this background, however, is increasing evidence of the lifelong plasticity of the brain and its responsiveness to experience (Kramer, Bherer, Colcombe, Dong, & Greenough, 004; Reuter-Lorenz, 00; Steven & Blakemore, 004). These effects have been reflected in structural changes consisting of increased cell density in regions responsible for the relevant cognitive experience, such as second-language learning (Mechelli et al., 004) or spatial navigation (Maguire et al., 000). Moreover, stimulating lifelong experiences have been shown to protect cognition against normal decline in older age as well as the onset of dementia (Stern, 00; Valenzuela & Sachdev, 006a, 006b). The promise of such protection motivates the search for the experiences that yield these effects. One such experience is bilingualism. In various studies, it has been shown that children (Carlson & Meltzoff, 008; Mezzacappa, 004), adults (Bialystok, 006; Colzato et al., 008; Costa, Hernández, & Sebastián-Gallés, Ellen Bialystok, Department of Psychology, York University, and Rotman Research Institute of Baycrest; Anne-Marie DePape, Department of Psychology, York University, and Department of Psychology, McMaster University. This research was funded by a grant from the Natural Sciences and Engineering Research Council, Canada, to Ellen Bialystok. We are grateful to Fergus Craik, Takako Fujioka, Bernhard Ross, and Laurel Trainor for helpful discussions and suggestions on the background and design of the study, and to Xiaojia Feng, Gigi Luk, and Mythili Viswanathan for technical assistance in design and analysis. Correspondence concerning this article should be addressed to Ellen Bialystok, Department of Psychology, York University, 4700 Keele Street, Toronto, ON M3J 1P3, Canada. ellenb@yorku.ca 008), and older adults (Bialystok, Craik, Klein, & Viswanathan, 004; Bialystok, Craik, & Ryan, 006) who are lifelong bilinguals and use both languages on a regular basis demonstrate higher levels of executive control than comparable monolinguals. These advantages have been shown to hold up and protect bilinguals against the onset of dementia, with symptoms appearing at a significantly later age than they do for monolinguals (Bialystok, Craik, & Freedman, 007). The mechanism responsible for these effects is unclear, but the advantage is likely rooted in the need for bilinguals to constantly manage attention to two active language systems. It is now widely accepted that both languages are activated when a bilingual is speaking one of them, even in a strongly monolingual context (Beauvillain & Grainger, 1987; Brysbaert, Van Dyck, & Van de Poel, 1999; Colomé, 001; Costa, 005; De Groot, Delmaar, & Lupker, 000; Dijkstra, Grainger, & Van Heuven, 1999; Duyck, Van Assche, Drieghe, & Hartsuiker, 007; Hermans, Bongaerts, De Bot, & Schreuder, 1998; Jared & Kroll, 001; Rodriguez- Fornells et al., 005; Van Hell & Dijkstra, 00), so fluent speech production normally requires that bilinguals control attention to the target language (Abutalebi & Green, 007; Green, 1998). Proposals for the mechanism within the executive function system that is responsible for these enhancements include active inhibition (Green, 1998), reactive inhibition (Colzato et al., 008), selective attention (Costa, 005), switching (Meuter & Allport, 1999), and some combination of these. If this attentional control to language were handled by general executive processes, then those processes would be broadly enhanced and more efficient for a wide range of tasks, including nonverbal ones. Because the precise mechanism responsible for the bilingual advantage has not been identified, it is difficult to connect this research to the emerging literature on experiential effects on cognition. A striking feature of the bilingual advantages in cognition is that they are found in domains far removed from language processing, suggesting 565

2 566 BIALYSTOK AND DEPAPE the role of domain-general mechanisms. Therefore, other experiences that engage executive control through requirements specific to that skill may also show these domain-general advantages. Identifying experiences with similar consequences on cognition will not only help to isolate the mechanism by which bilingualism has this effect but will also contribute to our understanding of how experience potentially reorganizes control processes in cognitive functioning. The criteria for such experiences are that they are massively practiced and that they engage frontal executive processes. One candidate experience in which to investigate the impact on cognition is musical expertise. The amount of time that musicians spend engaged in musical practice is considerable: Ericsson, Krampe, and Tesch-Römer (1993) reported that successful violinists in a performance class had accumulated an average of 10,000 hours of practice by the time they were 1 years old. Moreover, musical performance demands high levels of control through the need for selective attention and inhibition, switching, and updating and monitoring. These are the three primary components of executive function identified by Miyake and Shah (1999) and were shown to be distinct aspects of executive control (Miyake, Friedman, Emerson, Witzki, & Howerter, 000). Because each of them has a role in musical performance, it is plausible that extensive musical experience also enhances general executive functioning. Therefore, musicians meet both criteria for massive practice and the involvement of executive control, so musicians who practice regularly over many years might show cognitive benefits in executive control similar to those found for bilinguals. Although there are clear differences between the processing involved in language production and musical performance, the similarities at the level of executive control are sufficiently compelling to justify the hypothesis of an overall similarity between these groups. The possibility that music enhances cognition became prominent with the claims for the Mozart effect (Rauscher, Shaw, & Ky, 1993). However, the original claim that brief exposure to a Mozart sonata improves spatial cognition has been refuted by studies attributing the effect to the level of positive mood and arousal associated with the experimental condition but not the control conditions (Nantais & Schellenberg, 1999; Schellenberg, 006). In contrast, Schellenberg (003) proposed that formal musical training could have long-term cognitive benefits and elsewhere even predicted a more precise relation between musical training and enhanced executive control (Schellenberg, 006). More specific evidence for the effect of musical experience on cognition has also been documented. Schön, Magne, and Besson (004) asked participants to make pitch judgments about melodies and sentences in which the last musical note or spoken word was congruent, weakly incongruent, or highly incongruent with previous stimuli. Musicians were faster and more accurate than nonmusicians in all conditions for both types of stimuli. In related research, Ho, Cheung, and Chan (003) reported improvements in verbal memory, but not visual memory, for individuals taking music lessons; these effects persisted 1 year later for those who continued music lessons but disappeared when training was discontinued. More generally, when variables such as family income were controlled, Schellenberg (004) reported an increase in intelligence among children who had received either keyboard or voice lessons. In contrast, participants in a control condition who received drama classes showed an improvement in social functioning with no increase in IQ, demonstrating the specificity of some experiential effects. The mechanism by which musical experience might affect cognitive processing seems even less clear than that for language experience because the networks of activation are so diffuse (Altenmüller, 003). Nonetheless, there is evidence confirming the involvement of executive processes in musical performance, making a parallel case with bilingualism possible. Zatorre and colleagues (Zatorre, Evans, & Meyer, 1994) reported a PET study in which there was significant activation of regions in the right frontal cortex when comparing the pitch of the first and second tone of a melody and much greater frontal activation when comparing the first and last note, presumably because of the greater demands on working memory. Hall and Blasko (005) presented two tonal stimuli created by two instruments, violin or clarinet, to musicians with either the same instrument or different instruments presented to each ear. Participants needed to identify the sound in the target ear and ignore the sound in the other ear. Trials were congruent if the same instrument was heard in both ears or incongruent if there was a mismatch. The ability to respond to the incongruent trials increased with the number of years of musical experience. Similarly, Goolsby (1994a, 1994b) showed that musicians who were more skilled sight readers had better control over eye fixations in reading the music, demonstrating better ability to monitor and shift attention; specifically, skilled sight readers fixated on notes further ahead from those they were presently playing. Other evidence confirms the involvement of frontal regions, and therefore executive processes, in musical performance. Altenmüller (1989) presented melodic and harmonic discrimination tasks to participants who were classified as nonmusicians, amateur musicians, or professional musicians. Both amateur and nonmusicians processed the tasks in the right temporal and frontal lobes, whereas professional musicians processed them in the left frontotemporal lobes. More surprisingly, Ohnishi et al. (001) reported that musicians, but not nonmusicians, engaged frontal regions such as the left dorsolateral prefrontal cortex even in passive listening to music. There is little research connecting musical experience specifically to changes in these executive functions, especially in tasks that are not themselves auditory or musical, but one study with older adults is suggestive. Bugos and colleagues (Bugos, Perlstein, McCrae, Brophy, & Bedenbaugh, 007) trained older adults with individual piano instruction for 6 months and found some improvement in executive control, particularly in the ability to perform the trail making task and the digit span task relative to a control group. However, the study was based on a small sample, and the reported effects were modest. Some neuroimaging studies have shown that musical experience affects regions of the brain directly involved in musical performance. Elbert, Pantev, Wienbruch, Rockstroh, and Taub (1995) used magnetoencephalography and found that string musicians with approximately 1 years of experience had larger representational extent for the fingers in the left hand, the primary hand used by string players, than nonmusicians, with no difference between groups for the left thumb or for the fingers in the right hand. Furthermore, the amount of cortical enlargement was positively correlated with the number of years of musical experience. Subsequently, Pascual-Leone (003) replicated these results in a training study in which nonmusicians performed finger exercises over

3 MUSICAL EXPERTISE AND EXECUTIVE CONTROL 567 an extended period of time. After achieving near-perfect performance, half the participants stopped practicing and the other half continued to perform daily finger activities. Focal transcranial magnetic stimulation revealed that participants who stopped performing the finger exercises showed a return to baseline after 1 week, but those who continued practicing showed a significant positive correlation between the exercises and reorganization in the cortical motor outputs corresponding to the implicated fingers in this activity. These results indicate that changes occurring at the cortical level in musicians are more likely the result of extensive practice than a consequence of brain differences that are present before this activity is initiated. The possibility that musical experience has consequences similar to those found for language experience is also supported by evidence for the interrelatedness of music and language in the brain. Patel, Gibson, Ratner, Besson, and Holcomb (1998) showed that the ERP component elicited by a syntactic violation in a sentence or a harmonic violation in a musical phrase is the same. In a behavioral study, Poulin-Charronnat and colleagues (Poulin- Charronnat, Bigand, Madurell, & Peereman, 005) showed that harmonic structure interfered with performance on a lexical decision task, an effect found equally for musicians and nonmusicians. These results suggest an interconnection between the cognitive processes involved in language and music, making it plausible that extensive practice of either would have similar effects on both domains. Studies with musicians have typically investigated skilled instrumentalists. In addition to the demands of musical comprehension and production, instrumentalists face challenges to fine motor control; a skilled piano performance may include the production of 0 notes per second (Sloboda, 000). Therefore, if enhanced levels of executive processing were found for instrumentalists, it would be difficult to identify the source of that advantage in the musical or motoric dimensions of their experience. One way to isolate the source is to compare instrumentalists to vocalists who have similar musical experience but without the same type of motor demands that accompany instrumental playing. Singers certainly face considerable challenges to fine motor control, but the overt expression of that coordination is less obvious. Although there is no a priori reason to believe that vocalists develop different underlying cognitive skills from instrumentalists, the novelty of the research question makes it important to establish the boundaries around which such effects are observed. If musical expertise leads to enhanced executive control, then there should be no difference between those who play instruments and those who primarily spend their time singing. To investigate this issue, we included both instrumentalists and vocalists in the study design. In the present study we investigated whether intensive musical experience leads to enhancements in executive processing. The research with bilinguals showed that the effect of linguistic experience was not confined to linguistic outcomes but was found as well in nonlinguistic domains that relied on similar control processes. Therefore, the generality of the potential effect of musical experience was examined by manipulating the similarity of the tasks to musical expertise. If musical experience has an effect on executive functioning, then it will be most likely observed in a task requiring auditory judgments. Therefore, one task was an auditory adaptation of the Stroop task, the most commonly used task of executive control. If musical experience has a more generalized effect, as found for bilingualism, then the consequences would also be seen in nonverbal spatial tasks. Therefore, a spatial conflict task shown to be performed better by bilinguals, the Simon task, was included to determine the generality of these potential effects. The participants included bilinguals, to replicate the results found in previous research; monolingual musicians, to test these experiential effects with a new group; and monolinguals who were not musicians. Participants Method There were 95 participants ranging from 18 to 35 years old (M 3.8, SD 4.1). Monolingual speakers (N 4) spoke only English, and bilinguals (N 4) were fluent in English plus 1 of 14 other languages (French, Hebrew, Cantonese, Hindi, Punjabi, Spanish, Greek, Mandarin, Russian, Italian, Polish, Urdu, Gujarati, and Japanese). The bilinguals reported using English about 56% of the time each day and the other language in about 44% of daily activities. Bilinguals rated themselves on a 5-point scale as being highly fluent in English (M 4.5, SD 0.68) and moderately fluent in their other language (M 3.4, SD 0.93). The musician groups consisted of instrumentalists who played at least 1 of 13 instruments (piano, trumpet, violin, clarinet, guitar, drums, baritone, flute, xylophone, saxophone, cello, organ, and viola) and 5 vocalists who were classically trained. All the musicians were monolingual speakers of English. Tasks The background tasks were selected to establish the level of relevant experience for participants in each group (Language and Musical Background Questionnaire, which we created for this study), confirm that there was no difference in general intellectual level among the groups (Cattell Culture Fair Intelligence Test; Cattell & Cattell, 1960), establish comparable levels of spatial memory (Spatial Span Subtest; Wechsler, 1997), and demonstrate equivalent performance in a widely used neuropsychological measure of cognitive functioning (Trail Making Test; Lezak, 1995). The two experimental tasks that assessed the resolution of conflict were chosen to include a spatial task that had no obvious connection to either linguistic or musical experience and had previously been shown to be performed better by bilinguals (Simon task) and a task developed to present auditory conflict that had the same structure as the Simon task but was specific to auditory processing (Stroop task). These tasks were also used to determine the extent to which potential enhancements in executive control by musicians were domain general (Simon task), domain specific (Stroop task), or both. Language and Musical Background Questionnaire. This selfreport questionnaire recorded background and demographic information as well as experience with other languages and musical performance. Language fluency was rated on a 5-point scale ranging from not fluent (1) to extremely fluent (5). Participants were also asked to estimate the percentage of daily activities conducted in each language. Musical experience was assessed by the number of hours of daily practice, the method of musical instruction that

4 568 BIALYSTOK AND DEPAPE they received, and the context in which they engaged in this activity, such as recitals. Cattell Culture Fair Intelligence Test. This is a nonverbal measure of general intelligence that consists of 50 items that are administered in four subtests, each with a fixed time limit. The subtests include completing series of pictures, identifying similarity in a set, and choosing an item to match to a sample. Raw scores were converted to age-corrected general intelligence scores. Spatial Span Subtest of the Wechsler Memory Scale III. This is a test of short-term visual spatial memory. A 1-cm 7.5-cm platform containing 10 randomly positioned cubes is shown to participants. The experimenter points to increasingly long strings of blocks that the participant must repeat in the same order for the forward span and then in the reverse order for the backward span. There are two trials at each string length. Testing began with two-block strings and progressed to a maximum of nine blocks per trial but terminated when the participant made mistakes in both trials for a given level. Trail Making Test. In Part A, a sheet containing circles numbered from 1 through 5 is presented, and participants are required to connect the circles in sequence as quickly as possible without making errors and without lifting the pen. Part B is considered a test of executive functioning. Each circle contains either a number or a letter, and participants connect the numbers and letters in alternating sequence, such as 1-A, -B, and so on. The score is the length of time in seconds required to complete each part. Simon arrows task. This task was a modified version of the original paradigm used by Simon and Rudell (1967). The task was programmed in E-prime (Psychology Software Tools, 00) and presented on a Dell Inspiron 6000 laptop with a 30-cm 3-cm monitor. A mouse was attached to each side of the computer, providing a right and left response key. Participants sat with the index finger of each hand resting on one of the mouse keys. There were four conditions: direction control, position control, opposite, and conflict (consisting of congruent and incongruent trials). There was a set of 10 practice trials before each condition. The four conditions were presented in a counterbalanced order; each condition consisted of 96 trials in random order. In each condition, the trial began with a fixation cross in the center of the screen and was replaced by the stimulus, which was a directional arrow, after 50 ms. The stimuli were black arrows shown on a white background. Each arrow was 6.5 cm long and 0.5 cm wide, becoming 1.5 cm at the widest point of the arrowhead. The stimulus remained on the screen for 550 ms, and the next trial began immediately after the response was made. Participants were instructed to respond as quickly and accurately as possible. The direction control condition measured response speed. An arrow appeared in the center of the screen, and participants indicated the direction that the arrow was pointing by pressing the response key on the corresponding side of the monitor. Half of the arrows pointed in each direction. The position control condition measured the ability to rapidly detect the spatial position. Arrows were presented in the right- or left-hand side of the screen, and participants indicated the position in which the arrow appeared (irrespective of its direction) by pressing the appropriate response key. Because the arrows were pointing in the same direction as the position or the opposite direction, the trials could be classified as congruent or incongruent. There were 48 trials presented in each spatial position, half of which were congruent. The opposite condition was designed to measure response inhibition (Bunge, Dudukovic, Thomason, Vaidya, & Gabrieli, 00). An arrow appeared in the center of the monitor, and participants responded by indicating the direction opposite to the one in which the arrow was pointing, again by pressing the appropriate response key. In the conflict condition, the arrow appeared on one side of the display (instead of the center), and participants indicated the direction the arrow was pointing. Congruent trials were those for which the direction and the position were the same (e.g., a rightpointing arrow appearing on the right side of the screen); incongruent trials were those for which the values conflicted (e.g., a right-pointing arrow appearing on the left side of the screen). The Simon effect is the increase in reaction time needed to respond to incongruent trials. Auditory Stroop task. This measure was a modified version of the original task created by Hamers and Lambert (197). The task was programmed in E-prime (Psychology Software Tools, 00) and presented on the same laptop computer used for the Simon arrows task. Response keys were positioned on each side of the monitor. There were four conditions: pitch control, word control, pitch conflict, and word conflict. Each condition was preceded by 10 practice trials. The conditions were presented in counterbalanced order across participants. Each trial began with a fixation cross in the center of the screen that remained until a response was made. The auditory stimulus appeared 50 ms after the onset of the fixation and was presented for 550 ms. In the pitch control condition consisting of 48 trials, the stimuli were a high-pitched ( Hz) or low-pitched ( Hz) ahh recorded in Adobe Audition 1.5 (Adobe Systems Incorporated, 005). Half of the participants were instructed to indicate the pitch by pressing the right mouse button for high-pitched stimuli and left mouse button for lowpitched stimuli, and the other half were instructed to do the opposite. The letters H for high and L for low also appeared in the appropriate corners of the display to remind participants on which side to make their response. The word control condition was similar and also consisted of 48 trials. In this case, the stimuli were the spoken words high and low recorded in a female voice. Again, the response side corresponding to high and low was counterbalanced. In the conflict conditions, the same pitches used in the pitch control condition were produced using the words high or low instead of the syllable ahh. When pitch and word corresponded, that is, a high pitch sung with the word high, the trials were congruent; when pitch and words conflicted, the trials were incongruent. The stimuli were sung by the same female voice and recorded at Hz and Hz for high and low, respectively. There were 96 trials, of which half were congruent and half incongruent. In the pitch conflict condition, participants judged the pitch as high or low irrespective of the word; in the word conflict condition, they indicated the word irrespective of the pitch. Again, the assignment of high and low to the right and left keys was counterbalanced across participants. The entire task consisted of 576 trials including 19 control trials (48 4) and 384 conflict trials (96 4), which required approximately 15 min to complete.

5 MUSICAL EXPERTISE AND EXECUTIVE CONTROL 569 Results Table 1 shows the mean values and standard deviations for the demographic variables and background measures by participant group. There were no significant differences in participants ages across groups, F(3, 91) 1.44, ns. The groups differed in the mean number of years of musical experience, F(3, 91) 40., MSE 95.60, p.001, and amount of daily musical practice, F(3, 91).1, MSE 0.9, p.001, with the two musician groups differing from the nonmusicians, and no difference within each of these pairs (Scheffé, p.01). The groups differed as well in all the language background questions, with significant effects for English fluency, F(3, 91) 8.86, p.0001; non-english fluency, F(3, 91) 38.7, p.0001; and proportion of time speaking English in contrast to the non-english language, F(3, 91) , p In all cases, the bilinguals were different from the other three groups, with no differences among those three (Scheffé, p.01). It should be noted, however, that even though the bilingual measure for English fluency was significantly lower than that for the other three groups, it was still very high (4.3 out of 5) and clearly within the range of normal language proficiency. The mean scores on the background cognitive tasks are reported in Table. A series of one-way analyses of variance (ANOVAs) showed no significant differences between groups on any of these tasks, including the Cattell Culture Fair Intelligence Test, F(3, 87) 1.36, ns; Forward Spatial Span, F(3, 85).10, ns; Backward Spatial Span, F(3, 91).1, ns; Trail Making A, F(3, 90) 1.74, ns; and Trail Making B, F 1. Simon Arrows Task Mean accuracy rates for the Simon arrows task ranged between 93% and 99% and were equivalent across participant groups. The raw data were trimmed to exclude reaction times (RTs) that were less than 50 ms or greater than,500 ms. These long RTs, the equivalent of two standard deviations above the mean, were replaced with missing values. This replacement occurred for less than 0.1% of total responses (19 out of 31,90 responses). The mean RTs for the direction and position control conditions are presented in Figure 1. A two-way ANOVA for group and condition showed that the position control was faster than the direction control, F(1, 84) 8.76, MSE 3,648, partial.11, p.004, with no difference between groups, F(3, 84).46, ns, and no interaction, F 1. A one-way ANOVA on the position condition comparing the congruent and incongruent trials showed no difference between them, F 1. Results for the opposite and conflict conditions are presented in Figure. A two-way ANOVA for condition and group indicated a significant effect for condition, F(, 168) , MSE 139,770, partial.61, p.0001, with all conditions different from each other, ps.001, and group, F(3, 84) 4.59, MSE 3,056, partial.15, p.005, with monolinguals performing more slowly than the other three groups and no difference among the latter three, p.01. There was no interaction of condition and group, F 1. Auditory Stroop Task The mean accuracy rates ranged between 89% and 99% with no difference across groups. The raw data were trimmed and reviewed for outliers using the same procedure described in the Simon arrows task. In total, missing values accounted for less than 0.1% of total responses ( out of 54,70 responses). The results for the two control conditions, presented in Figure 3, were examined with a three-way ANOVA for condition (word, pitch), mapping (right, left), and group. The pitch condition elicited faster RTs than the word condition, F(1, 78) , MSE 415,691, partial.65, p.0001, and mapping high onto the right side produced faster responses than mapping it onto the left, F(1, 78) 13.7, MSE 5,96, partial.15, p There was no difference across groups, F(3, 78) 1.64, ns, and no interaction. The musicians appeared to be responding faster on the pitch control condition than the other groups, but a one-way ANOVA for RT in the pitch control condition was not quite significant, F(3, 8).37, p.08. The mean reaction times for the pitch conflict condition are shown in Figure 4. A two-way ANOVA indicated main effects for group, F(3, 88) 7.63, MSE 01,104, partial.1, p.001; congruency, F(1, 88) 75.49, MSE 143,40, partial.46, p.001; and their interaction, F(3, 88).86, MSE 5,47, partial.09, p.05. Contrasts showed that the monolinguals and bilinguals were both slower than the instrumentalists and vocalists, with no difference within each of these pairs. The group difference was found for both congruent and incongruent trials, although the magnitude of that difference was somewhat larger for incongruent Table 1 Mean Values and Standard Deviations for the Demographic Variables by Participant Group Language group Music group Monolingual Bilingual Instrumental Vocalist Variable M SD M SD M SD M SD Age (years) Musical experience (years) Daily musical practice (hr) Fluency in English (max. 5) Fluency in non-english (max. 5) Proportion English spoken daily Note. max. maximum.

6 570 BIALYSTOK AND DEPAPE Table Mean Scores and Standard Deviations on Background Cognitive Measures by Participant Group Language group Music group Monolingual Bilingual Instrumental Vocalist Measure M SD M SD M SD M SD Cattell Forward Span Backward Span Trails A (s) Trails B (s) trials, F(3, 88) 8.5, p.0001, than for the congruent trials, F(3, 90) 6.00, p A comparison of the RT cost difference between congruent and incongruent trials (Stroop effect) across groups showed that the smallest cost was for the instrumentalists, who took 9 ms extra for incongruent trials; the largest for the nonmusicians, with monolinguals taking 71 ms and bilinguals taking 76 ms; and the vocalists in between with 48 ms. Although this difference between groups was significant, F(3, 88).86, p.04, contrasts failed to reveal significant pairwise differences between groups. Although the group difference in RT to the pitch control condition was not significant, the groups appeared to be different. Therefore, a new analysis that was based on individual RT in the control condition was conducted to verify the group differences in the experimental task. The proportional cost of the Stroop effect was determined for each participant by calculating the ratio of the Stroop effect (difference between congruent and incongruent RT) to the control RT. Thus, RT differences in the control trials were irrelevant as the data compare only the additional cost to those times for the experimental trials. The proportion increases over the control trials for the Stroop effect were 0.15 for the monolinguals, 0.15 for the bilinguals, 0.06 for the instrumentalists, and 0.09 for the vocalists, a difference that was significant, F(3, 80).90, p.04, with differences between the musicians and the other groups. Therefore, the small difference in speed in the control pitch condition did not account for the performance differences on the conflict trials of this task. A two-way ANOVA was applied to the results of the word conflict condition, shown in Figure 5. There was an effect of congruency, F(1, 90) 50.8, MSE 59,484, partial.36, p.0001; no effect of group, F 1; but an interaction of group and congruency, F(3, 90) 3.01, p.04. To explore this interaction, we calculated the Stroop effect for each group as the difference between congruent and incongruent trials. Unlike the results for the pitch conflict, the smallest conflict in this case was found for the bilinguals, 18 ms, and the longest for instrumentalists, 60 ms, with the monolinguals, 9 ms, and vocalists, 33 ms, in between. A one-way ANOVA showed that the difference was significant, F(3, 90) 3.01, p.03, although the only significant contrast was between the bilinguals and the instrumentalists. Discussion Participants were selected for inclusion in the study if they were (a) lifelong bilinguals who regularly used both languages, (b) musicians who had been studying either an instrument or voice for at least half of their lives, or (c) monolingual nonmusicians who had neither specialized experience. In addition, the bilinguals were not musicians and the musicians were all monolingual, so the design isolated the effect of each of these experiences. All the participants were the same age and performed the same background measures assessing general intelligence (Cattell), speed and cognitive processing (Trails), and working memory (spatial span). They differed, however, in their performance on the executive control tasks. On a nonverbal spatial conflict task, the bilinguals and the musicians outperformed the monolinguals; on an auditory conflict task, the musicians outperformed both the monolinguals and the bilinguals on overall response speed but only Mean RT (ms) Direction Position Condition Mean RT (ms) Opposite Congruent Incongruent Condition Monolingual Bilingual Instrument Vocal Monolingual Bilingual Instrument Vocal Figure 1. Mean reaction times (RT) and standard errors for the direction control and position control (collapsed across congruency) conditions in the Simon arrows task. Figure. Mean reaction times (RT) and standard errors for the opposite condition, and congruent and incongruent trials of the conflict condition in the Simon arrows task.

7 MUSICAL EXPERTISE AND EXECUTIVE CONTROL Mean RT (ms) Mean RT (ms) P i t c h W o r d Condition Monolingual Bilingual Instrument Vocal 450 C o n g r u e n t I n c o n g r u e n t Condition Monolingual Bilingual Instrument Vocal Figure 3. Mean reaction times (RT) and standard errors for the combined pitch control and word control conditions in the auditory Stroop task. showed a smaller incongruent cost for conditions based on pitch conflict and not on word conflict. The Simon arrows task was an adaptation of nonverbal conflict tasks used in our previous research, and the auditory Stroop task was based on a task developed by Hamers and Lambert (197). Both tasks included simple control conditions that allowed comparisons across groups for initial differences in response speed. For the Simon arrows task, these included the direction control, in which participants indicated the direction of the arrow in the absence of position information, and the position control, in which participants indicated the location of the stimulus irrespective of the direction it was pointing. For the auditory Stroop task, these included the pitch control and word control conditions, in which judgments of high and low were made in the absence of competing or facilitating information from the pitch or the word. In both cases, all the participants performed at the same speed and with the same accuracy on these control conditions. Therefore, there is no evidence that participants in the four groups approached these tasks with different levels of expertise in the simplest versions. The experimental conditions provide evidence for both the specific executive components involved in these tasks and the effect of these two experiences on those components. In these conflict conditions, the performance of the four groups diverged. Consider first the Simon arrows task. In the opposite condition, participants were required to override a prepotent response and indicate the direction contrary to that shown by the arrow; in the conflict conditions, participants were required to monitor the direction of the arrow irrespective of its position on the display in order to attend to a target in the context of misleading information. Mean RT (ms) C o n g r u e n t Condition I n c o n g r u e n t Monolingual Bilingual Instrument Vocal Figure 4. Mean reaction times (RT) and standard errors for the pitch conflict conditions by congruent and incongruent trials in the auditory Stroop task. Figure 5. Mean reaction times (RT) and standard errors for the word conflict condition by congruent and incongruent trials in the auditory Stroop task. These abilities are both aspects of executive control, although they engage different components (Bunge et al., 00). In both cases, however, the monolinguals performed more slowly than the bilinguals and musicians, with no difference between these latter groups. Thus, the two experiences were equally effective in boosting each of these processing components. Although other studies have revealed bilingual benefits in conflict but not opposite condition versions of similar problems (Bialystok, Craik, & Ryan, 006; Carlson & Meltzoff, 008; Martin-Rhee & Bialystok, 008), the present results showed all executive control components to be enhanced through one of these experiences. Research investigating the effect of bilingualism on executive control consistently reports advantages in the conflict conditions, but performance in the withholding conditions is less consistent. More research is needed to investigate the conditions under which these effects occur for response inhibition. In both the Simon and Stroop tasks, participants with special experiences performed more rapidly than monolinguals on both congruent and incongruent trials of the conflict conditions. From the hypothesis of enhanced executive control, it might have been predicted that the effect would be found only on the more difficult incongruent trials where selective attention or inhibition was more obviously required. However, in previous research with such paradigms, enhanced performance has been reported for both congruent and incongruent trials when they are presented randomly in a mixed block (e.g., Bialystok, 006; Bialystok, Craik, Klein, & Viswanathan, 004). Under these conditions, performance also requires monitoring and switching, these also being components of executive control (Miyake & Shah, 1999). The implication is that monitoring and switching are two of the processes boosted by experiences that enhance executive control. In a previous study using a different paradigm but a structurally similar design, monolinguals and bilinguals responded to congruent and incongruent trials of a behavioral antisaccade task. In mixed block presentation, bilinguals performed faster than monolinguals on both congruent and incongruent trials as in the present study, but in a block consisting of only congruent trials, there was no group difference (Bialystok, Craik, & Ryan, 006). This pattern is consistent with the interpretation that the faster performance on the congruent trials by the bilingual and musician groups in the present study is attributable to the executive control costs for monitoring and switching associated with mixed block presentations. The auditory Stroop task included two kinds of conflict trials because it was not known which type of interference would be

8 57 BIALYSTOK AND DEPAPE most problematic and whether the different experiences would have different impacts on each. In the word conflict condition in which participants indicated the word irrespective of pitch, the cost of the incongruent trials compared with the congruent trials was greatest for the instrumentalists and smallest for the bilinguals, with the other groups in between; the vocalists performed more like the bilinguals than like the instrumentalists. In the pitch conflict condition, in contrast, participants in the two musician groups performed faster than the monolinguals and the bilinguals and also recorded smaller costs for the incongruent trials. Thus, even within an auditory task, the locus of the conflict was associated with enhanced performance by those for whom that feature was part of their experience. In both experimental tasks, participants in the two musician groups performed similarly to each other and differently from those in the other groups. Despite careful attention to the background of the participants, however, it turned out that many of the vocalists had extensive instrumental training, usually on piano. Therefore, it is possible that the enhancing effect of musical experience is attributable to instrumental practice rather than vocal training. To check that possibility, we divided the vocalists into subgroups according to whether they had instrumental training. This process identified 10 instrumental vocalists who had received an average of 10.6 years of instrumental training and 15 vocalists with no such training. These subgroups were compared across the experimental tasks, and none of the analyses was significant. Therefore, the nature of the musical experience in terms of instrumental or vocal training made no difference to the outcome observed for performance on executive control tasks. The equivalent effects for musicians with two different types of training increase confidence that it is musical expertise in general that was responsible for the outcomes. The results of the present study demonstrate the effectiveness of musical training and musical practice on enhancing executive control, even for tasks that bear no obvious relation to music. Even with comparable performance on the background and control tasks, musicians and bilinguals performed similarly to each other and were both superior to monolinguals with no musical experience on a spatial conflict task. On a task more situated in an auditory experience, however, the two musician groups outperformed the other participants, including the bilinguals, in that their response times were faster and their ability to resolve conflict was better if that conflict was based on pitch. This pattern indicates that the effect of experience can generalize to other domains but that the greatest effect is found in tasks that are similar to the type of activity involved in the experience itself. Thus, the study provides evidence connecting musical experience to executive control, showing both domain-specific effects in an auditory task (auditory Stroop) and domain-general effects in a spatial task (Simon arrows). It also extends the research on executive control in bilinguals by showing that such enhancement is not specific to language experience and opens the possibility that other types of experience in highly controlled processing may also lead to these effects. Despite clear behavioral results for the enhancement of executive control in musicians and bilinguals, the mechanism underlying these effects is only speculative. In the case of bilingualism, there is reason to believe that the enhanced executive control follows from the constant use of these processes to monitor attention to two languages. Musical performance is also a complex task and engages selective attention, monitoring, and shifting, processes that may well underlie the effects demonstrated here. However, more research is needed to establish clearer links between the involvement of these processes in musical performance and their use in the tasks examined in the present study. This research also contributes to the identification of other experiences that may have the potential to build up levels of executive functioning; the present results suggest that at least two of the criteria for these experiences are that they are massively practiced and engage executive control in their execution. Even without clear evidence for the underlying mechanism, the results add to a growing body of research demonstrating the important role of experience in shaping general cognitive processes. References Abutalebi, J., & Green, D. (007). Bilingual language production: The neurocognition of language representation and control. Journal of Neurolinguistics, 0, Adobe Systems Incorporated. (005). Adobe audition 1.5. San Jose, CA: Adobe Systems Incorporated. Altenmüller, E. (1989). Cortical DC-potentials as electrophysiological correlates of hemispheric dominance of higher cognitive functions. International Journal of Neuroscience, 47, Altenmüller, E. O. (003). How many music centers are in the brain? In I. Peretz & R. Zatorre (Eds.), The cognitive neuroscience of music (pp ). Oxford, England: Oxford University Press. Beauvillain, C., & Grainger, J. (1987). Accessing interlexical homographs: Some limitations of a language-selective access. Journal of Memory and Language, 6, Bialystok, E. (006). Effect of bilingualism and computer video game experience on the Simon task. Canadian Journal of Experimental Psychology, 60, Bialystok, E., Craik, F. I. M., & Freedman, M. (007). Bilingualism as a protection against the onset of symptoms of dementia. Neuropsychologia, 45, Bialystok, E., Craik, F. I. M., Klein, R., & Viswanathan, M. (004). Bilingualism, aging, and cognitive control: Evidence from the Simon task. Psychology and Aging, 19, Bialystok, E., Craik, F. I. M., & Ryan, J. (006). Executive control in a modified antisaccade task: Effects of aging and bilingualism. Journal of Experimental Psychology, 3, Brysbaert, M., Van Dyck, G., & Van de Poel, M. (1999). Visual word recognition in bilinguals: Evidence from masked phonological priming. Journal of Experimental Psychology: Human Perception and Performance, 5, Buckner, R. L., Head, D., & Lustig, C. (006). Brain changes in aging: A lifespan perspective. In E. Bialystok & F. I. M. Craik (Eds.), Lifespan cognition: Mechanisms of change (pp. 7 4). New York: Oxford University Press. Bugos, J. A., Perlstein, W. M., McCrae, C. S., Brophy, T. S., & Bedenbaugh, P. H. (007). Individualized piano instruction enhances executive functioning and working memory in older adults. Aging & Mental Health, 11, Bunge, S. A., Dudukovic, N. M., Thomason, M. E., Vaidya, C. J., & Gabrieli, J. D. E. (00). Immature frontal lobe contributions to cognitive control in children: Evidence from fmri. Neuron, 33, Carlson, S. M., & Meltzoff, A. M. (008). Bilingual experience and executive functioning in young children. Developmental Science, 11, Cattell, R. B., & Cattell, A. K. S. (1960). The individual or group Culture Fair Intelligence Test. Champaign, IL: Institute for Personality and Ability Testing.

9 MUSICAL EXPERTISE AND EXECUTIVE CONTROL 573 Colomé, À. (001). Lexical activation in bilinguals speech production: Language-specific or language-independent? Journal of Memory and Language, 45, Colzato, L., Bajo, M. T., van den Wildenberg, W., Paolieri, D., Nieuwenhuis, S., La Heij, W., et al. (008). How does bilingualism improve executive control? A comparison of active and reactive inhibition mechanisms. Journal of Experimental Psychology: Learning, Memory, and Cognition, 34, Costa, A. (005). Lexical access in bilingual production. In J. F. Kroll & A. M. B. de Groot (Eds.), Handbook of bilingualism: Psycholinguistic approaches (pp ). Oxford, England: Oxford University Press. Costa, A., Hernández, M., & Sebastián-Gallés, N. (008). Bilingualism aids conflict resolution: Evidence from the ANT task. Cognition, 106, De Groot, A. M. B., Delmaar, P., & Lupker, S. J. (000). The processing of interlexical homographs in translation recognition and lexical decision: Support for non-selective access to bilingual memory. The Quarterly Journal of Experimental Psychology A: Human Experimental Psychology, 53, Dijkstra, T., Grainger, J., & van Heuven, W. J. B. (1999). Recognition of cognates and interlingual homographs: The neglected role of phonology. Journal of Memory and Language, 41, Duyck, W., Van Assche, E., Drieghe, D., & Hartsuiker, R. J. (007). Visual word recognition by bilinguals in a sentence context: Evidence for nonselective lexical access. Journal of Experimental Psychology: Learning, Memory, and Cognition, 33, Elbert, T., Pantev, C., Wienbruch, C., Rockstroh, B., & Taub, E. (1995). Increased cortical representation of fingers of the left hand in string players. Science, 70, Ericsson, K. A., Krampe, R. T., & Tesch-Römer, C. (1993). The role of deliberate practice in the acquisition of expert performance. Psychological Review, 100, Goolsby, T. W. (1994a). Eye movement in music reading: Effects of reading ability, notational complexity, and encounters. Music Perception, 1, Goolsby, T. W. (1994b). Profiles of processing: Eye movements during sightreading. Music Perception, 1, Green, D. W. (1998). Mental control of the bilingual lexico-semantic system. Bilingualism: Language and Cognition, 1, Hall, M. D., & Blasko, D. G. (005). Attentional interference in judgments of musical timbre: Individual differences in working memory. The Journal of General Psychology, 13, Hamers, J. F., & Lambert, W. (197). Bilingual interdependencies in auditory perception. Journal of Verbal Learning and Verbal Behaviour, 11, Hermans, D. T., Bongaerts, T., de Bot, K., & Schreuder, R. (1998). Producing words in a foreign language: Can speakers prevent interference from their first language? Bilingualism: Language and Cognition, 1, Ho, Y-C., Cheung, M-C., & Chan, A. S. (003). Music training improves verbal but not visual memory: Cross-sectional and longitudinal explorations in children. Neuropsychology, 17, Jared, D., & Kroll, J. F. (001). Do bilinguals activate phonological representations in one or both of their languages when naming words? Journal of Memory and Language, 44, 31. Kramer, A. F., Bherer, L., Colcombe, S. J., Dong, W., & Greenough, W. T. (004). Environmental influences on cognitive and brain plasticity during aging. Journal of Gerontology: Medical Sciences, 59A, Lezak, M. D. (1995). Neuropsychological assessment (3rd ed.). New York: Oxford University Press. Maguire, E. A., Gadian, D. G., Johnsrude, I. S., Good, C. D., Ashburner, J., Frackowiak, R. S., et al. (000). Navigation-related structural changes in the hippocampi of taxi drivers. Proceedings of the National Academy of Sciences, 97, Martin-Rhee, M. M., & Bialystok, E. (008). The development of two types of inhibitory control in monolingual and bilingual children. Bilingualism: Language and Cognition, 11, Mechelli, A., Crinion, J. T., Noppeney, U., O Doherty, J., Ashburner, J., Frackowiak, R. S., et al. (004). Structural plasticity in the bilingual brain. Nature, 431, 757. Meuter, R. F. I., & Allport, A. (1999). Bilingual language switching in naming: Asymmetrical costs of language selection. Journal of Memory and Language, 40, Mezzacappa, E. (004). Alerting, orienting, and executive attention: Developmental properties and sociodemographic correlates in an epidemiological sample of young, urban children. Child Development, 75, Miyake, A., Friedman, N. P., Emerson, M. J., Witzki, A. H., & Howerter, A. (000). The unity and diversity of executive functions and their contributions to complex frontal lobe tasks: A latent variable analysis. Cognitive Psychology, 41, Miyake, A., & Shah, P. (1999). Toward unified theories of working memory: Emerging general consensus, unresolved theoretical issues, and future research directions. In A. Miyake & P. Shah (Eds.), Models of working memory: Mechanisms of active maintenance and executive control (pp ). New York: Cambridge University Press. Nantais, K. M., & Schellenberg, E. G. (1999). The Mozart effect: An artifact of preference. Psychological Science, 10, Ohnishi, T., Matsuda, H., Asada, T., Aruga, M., Hirakata, M., Nishikawa, M., et al. (001). Functional anatomy of musical perception in musicians. Cerebral Cortex, 11, Pascual-Leone, A. (003). The brain that makes music and is changed by it. In I. Peretz & R. Zatorre (Eds.), The cognitive neuroscience of music (pp ). Oxford, England: Oxford University Press. Patel, A. D., Gibson, E., Ratner, J., Besson, M., & Holcomb, P. (1998). Processing syntactic relations in language and music: An event-related potential study. Journal of Cognitive Neuroscience, 10, Poulin-Charronnat, B., Bigand, E., Madurell, F., & Peereman, R. (005). Musical structure modulates semantic priming in vocal music. Cognition, 94, B67 B78. Psychology Software Tools. (00). E-Prime 1.1. Pittsburgh, PA: University of Pittsburgh. Rauscher, F. H., Shaw, G. L., & Ky, K. N. (1993). Music and spatial task performance. Nature, 365, 611. Reuter-Lorenz, P. A. (00). New visions of the aging mind and brain. Trends in Cognitive Sciences, 6, Rodriguez-Fornells, A., van der Lugt, A., Rotte, M., Britti, B., Heinze, H-J., & Münte, T. F. (005). Second language interferes with word production in fluent bilinguals: Brain potential and functional imaging evidence. Journal of Cognitive Neuroscience, 17, Schellenberg, E. G. (003). Does exposure to music have beneficial side effects? In I. Peretz & R. J. Zatorre (Eds.), The cognitive neuroscience of music (pp ). Oxford, England: Oxford University Press. Schellenberg, E. G. (004). Music lessons enhance IQ. Psychological Science, 15, Schellenberg, E. G. (006). Long-term positive associations between music lessons and IQ. Journal of Educational Psychology, 98, Schön, D., Magne, C., & Besson, M. (004). The music of speech: Music training facilitates pitch processing in both music and language. Psychophysiology, 41, Simon, J. R., & Rudell, A. P. (1967). Auditory S-R compatibility: The effect of an irrelevant cue on information processing. Journal of Applied Psychology, 51, Sloboda, J. A. (000). Individual differences in musical performance. Trends in Cognitive Science, 4, Stern, Y. (00). What is cognitive reserve? Theory and research application of the reserve concept. Journal of the International Neuropsychological Society, 8, Steven, M. S., & Blakemore, C. (004). Cortical plasticity in the adult