music training facilitates lexical stress processing

Music Perception VOLUME 26, ISSUE 3, PP. 235–246, ISSN 0730-7829, ELECTRONIC ISSN 1533-8312 © 2009 BY THE REGENTS OF THE UNIVERSITY OF CALIFORNIA. ALL

RIGHTS RESERVED. PLEASE DIRECT ALL REQUESTS FOR PERMISSION TO PHOTOCOPY OR REPRODUCE ARTICLE CONTENT THROUGH THE UNIVERSITY OF CALIFORNIA PRESS’S

RIGHTS AND PERMISSIONS WEBSITE, HTTP://WWW.UCPRESSJOURNALS.COM/REPRINTINFO.ASP. DOI:10.1525/MP.2009.26.3.235

Music Training Facilitates Lexical Stress Processing 235

RÉGINE KOLINSKY, HÉLÈNE CUVELIER,AND VINCENT GOETRY

Université Libre de Bruxelles (ULB), Brussels, Belgium

ISABELLE PERETZ

Université de Montréal, Montréal, Canada

JOSÉ MORAIS

Université Libre de Bruxelles (ULB), Brussels, Belgium

WE INVESTIGATED WHETHER MUSIC TRAINING facili-tates the processing of lexical stress in natives of a lan-guage that does not use lexical stress contrasts.Musically trained (musicians) or untrained (nonmusi-cians) French natives were presented with two tasks:speeded classification that required them to focus on asegmental contrast and ignore irrelevant stress varia-tions, and sequence repetition involving either segmen-tal or stress contrasts. In the latter situation, Frenchnatives are usually “deaf” to lexical stress, but this wasless the case for musicians, demonstrating that musicexpertise enhances sensitivity to stress contrasts. Thisincreased sensitivity does not seem, however, tounavoidably bias musicians’ attention to stress con-trasts: in segmental-based speeded classification, musi-cians were not more affected than nonmusicians byirrelevant stress variations when overall performancewas controlled for. Implications regarding both thenotion of modularity of processing and the advantagethat musicianship may afford for second languagelearning are discussed.

Received May 22, 2008, accepted October 25, 2008.

Key words: lexical stress, stress deafness, musicianship,domain-transfer effects, modularity

INCREASING EVIDENCE POINTS TO A LINK BETWEEN

music lessons and prosodic skills in speech. Speechprosody refers to the “musical aspects” of speech,

including its “melody” (commonly called intonation)and its rhythm (stress and timing). In music, pitch andtemporal relations define musical tunes. In speech,duration, pitch height, intensity, and spectral variations

are to different degrees responsible for prosody, whichcontributes to speech communication through accents,pauses, and intonation that convey important sourcesof linguistic and emotional information.

Music training has been shown to enhance the abilityto decode emotions conveyed by speech prosody (e.g.,Nilsonne & Sundberg, 1985; Thompson, Schellenberg,& Husain, 2004). Indeed, emotions are also at the coreof music. However, this effect may be mediated by emo-tional intelligence rather than by music training per se(Trimmer & Cuddy, 2008). Prosody also providesindexical or extralinguistic information such as the talk-er’s voice, gender, or dialect, as well as important lin-guistic information, for example, allowing one todifferentiate questions from statements. More gener-ally, linguistic prosody helps listeners to determineboundaries between words and phrases. Descendingpitch contours and syllables of longer duration typi-cally mark ends of words or phrases in speech (e.g.,Beckman & Pierrehumbert, 1986; Klatt, 1976), suchas pitch drops and durational lengthening at the endof musical phrases (e.g., Narmour, 1990; Todd, 1985).Variations in pitch, duration, and intensity give cluesfor differentiating among different performances ofthe same music (Palmer, 1997; Palmer, Jungers, &Jusczyk, 2001) as well as among structural ambigui-ties (such as phrasal and metrical boundaries) thatarise in music (Palmer, 1996; Sloboda, 1985), just asprosodic features of speech can clarify the intendedmeaning of a syntactically ambiguous sentence(Lehiste, Olive, & Streeter, 1976; Price, Ostendorf,Shattuck-Hufnagel, & Fong, 1991; Scott, 1982).Moreover, both music and speech present accentswhose function is to capture the listener’s attentionduring the unfolding of the verbal message or of amusical piece (e.g., Jones & Boltz, 1989). This commonfunction is well illustrated by the fact that harmonicaccent allows faster processing of target phonemes inphoneme monitoring (Bigand, Tillmann, Poulin,D’Adamo, & Madurell, 2001).

Given these similarities, one may expect processing ofsubtle and hence difficult to detect prosodic variationsto benefit from music training. As a matter of fact, musi-cally trained adults and children (musicians) were foundto outperform untrained participants (nonmusicians) in

MUSIC TRAINING FACILITATES LEXICAL STRESS PROCESSING

Music2603_07 1/16/09 9:48 AM Page 235

detecting weak pitch violations in both melodies andintonation contours of spoken sentences (Magne,Schön, & Besson, 2006; Schön, Magne, & Besson, 2004).Phonetic experts also benefit from music training indifficult prosodic analyses such as identifying localvariations in nuclear linguistic tones (Dancovicová,House, Crooks, & Jones, 2007).

Languages also can be characterized by prosodic cuesassociated with lexical stress or tones. In tone lan-guages, some words are distinguished by the sole varia-tions in pitch contour. In comparison to nonmusicians,musicians have been shown to display more robustbrainstem encoding of Mandarin Chinese lexical tones(Wong, Skoe, Russo, Dees, & Kraus, 2007). Thus, via thecorticofugal pathway, music training can enhance audi-tory functions subserved by the rostral brainstem thatare relevant to speech.

Surprisingly, until now relatively little attention hasbeen paid to the relations between music training andthe processing of word stress patterns. Only some com-monalities between music structure and linguisticstress patterns have been reported. The prosody of acomposer’s native language influences the structure ofhis/her instrumental music (e.g., Abraham, 1974;Wenk, 1987). In English, which presents contrastiveopposition between stressed and unstressed syllables,these stress patterns tend to coincide with the metricalstructure in music (Palmer & Kelly, 1992; see alsoLerdahl, 2001). And, analyzing the instrumental musicof 16 classic British and French composers (e.g., Deliusvs. Saint-Saëns), Patel and Daniele (2003a, 2003b; seealso Huron & Ollen, 2003) observed the structure oftheir musical themes to parallel the one of their mothertongue, mainly by the greater variability of the tempo-ral patterning of vowels in (British) English than inFrench (Grabe & Low, 2002; Low, Grabe, & Nolan,2000; Ramus, Nespor, & Mehler, 1999).

In the present study, we explored the relationsbetween music and linguistic stress patterns in a differ-ent way: we investigated whether music training mightfacilitate the processing of lexical stress patterns innative speakers of French, a fixed stress language. Thislanguage does not use lexical stress contrasts: it displaysfixed word-final stress (e.g., Garde, 1968). French dif-fers from varying stress languages such as English,Dutch, and Spanish, which use stress to distinguish lex-ical items, allowing, for example, Spanish speakers tocontrast between BEbe1 (“drink”) and beBE (“baby”).

Lexical stress is instantiated by a variety of prosodiccues, namely pitch, energy, and duration, the impor-tance of which varies across languages (e.g., Garde,1968; Lehiste, 1970). In English, stressed syllables arehigher-pitched, louder and somewhat longer thanunstressed ones. In fact, English has two kinds ofunstressed vowels: reduced and full. Indeed, in this lan-guage (but not in others such as Spanish), the varia-tions of pitch, loudness and duration may beaccompanied by a modification of the quality of theunstressed vowel, which is then reduced to a central“schwa” (/∂/). For example, the quality of theunstressed first vowel of the verb obJECT /∂b’d εkt/ isquite different from that of the stressed first vowel ofthe noun OBject, /’ bd εkt/. In other minimal pairs,vocalic reduction is not as strong, such as in TRUSty vs.trusTEE. Within such pairs, stressed and unstressed syl-lables differ in vowel length, intensity, and pitch, but farless in vocalic quality.

To what extent are natives of languages such asFrench “deaf” to such lexical stress differences? Sincelexical stress contrasts have substantial acoustic corre-lates in terms of duration, fundamental frequency (F0),and energy, French natives are not expected to haveproblems with the perception of stress. In fact, Dupoux,Pallier, Sebastián, and Mehler (1997) reported that in asame/different task, French participants had no diffi-culty discriminating nonwords that differ only in theposition of stress, such as BOpelo-boPElo. However,French listeners were found to be “deaf” to stress con-trasts at a more abstract processing level.

Dupoux et al. (1997) used a speeded ABX task inwhich participants heard three successive items(VAsuma-vaSUma-VAsuma) and judged whether thethird item was identical to the first or the second one.This task requires a short-term memory buffer becausethe decision has to be delayed until the final stimulus isheard, and the phonetic variability linked to the factthat different talkers pronounced the three stimulimade an acoustically based response strategy more dif-ficult to use. Under these conditions, French nativesmade many more errors and were far slower thanSpanish natives. In addition, when they did not knowwhat dimension they had to pay attention to, they per-formed worse on stress (FIdape-fiDApe) than on seg-mental ( fidape-lidape) contrasts, and did not benefit,contrary to Spanish natives, from redundant stressinformation when segmental information alone wassufficient to perform the task (in FIdape-liDApe com-pared to fidape-lidape). Thus, only natives of a stresslanguage represent and jointly use both stress and seg-mental information.

236 Régine Kolinsky, Hélène Cuvelier, Vincent Goetry, Isabelle Peretz, & José Morais

1In all the examples, uppercase denotes stressed syllables, lower-case unstressed ones.

Music2603_07 1/16/09 9:48 AM 36

Peperkamp, Dupoux, and Sebastián-Gallés (1999)and Dupoux, Peperkamp, and Sebastián-Gallés (2001)observed even more robust stress deafness in Frenchnatives by using a sequence repetition task in whichperformance on a stress contrast was compared withthat on a control segmental contrast across differentlevels of memory load. The participants were firstrequired to associate two nonwords of a minimal pairdiffering only in one phonological dimension, eitherstress location (PIki–piKI) or segmental contrast(TUku–TUpu), with two different keyboard keys. Then,they listened to longer and longer random sequences ofthe two items, which they were required to recall andtranscribe as sequences of key presses. Analysis of thedifference scores (performance on stress contrast minusperformance on segmental contrast) revealed not onlya French-Spanish difference but also nonoverlappingdistributions of individual results between groups. Inaddition, the advantage of segmental over stress con-trast in French natives was found to be almost inde-pendent of sequence length. Interestingly, French nativeswho are late learners of Spanish show the same limita-tion in short-term memory encoding of stress contraststhat was found in French monolinguals (Dupoux,Sebastián-Gallés, Navarrete, & Peperkamp, 2008).

Other studies showed that only speakers of lan-guages in which stress is contrastive are unable toignore stress information, even in tasks in which thisfiltering capacity would be beneficial. Dupoux et al.(1997) found that Spanish natives hardly performed asegmental-based ABX task when stress variations wereput in conflict with the expected response (e.g., toanswer “A” to the ABX triple BOpelo–soPElo–boPElo), whereas French natives had no difficultyignoring the irrelevant stress variations. Using anadaptation of Garner’s (1974) speeded classificationtask, Pallier, Cutler, and Sebastián-Gallés (1997)showed that natives of both Spanish and Dutch (twovarying stress languages) had difficulties focusing onthe segments by filtering out irrelevant stress varia-tions in a segmental-based classification task thatrequired them to press one of two buttons correspon-ding respectively to deki and nusa, irrespective of stressposition. Indeed, both groups presented an interferenceeffect, namely longer reaction times (RTs) when thestress of the nonwords varied from trial to trial (DEkiand deKI associated to one response button vs. NUsaand nuSA to the other than when it was fixed (DEki vs.NUsa). The effect was larger for Spanish than forDutch, suggesting that the degree of interference fromstress variations may be mitigated by the predictabilityof stress, which is greater in Dutch than in Spanish.2

Here, we focused on the facilitation possibly inducedby music training on the processing of lexical stress pat-terns in native speakers of a language that does not uselexical stress contrasts. French natives, musicians, andnonmusicians were examined. Both groups were pre-sented with two tasks: an adaptation of the sequencerepetition task designed by Dupoux et al. (2001), inwhich they had to process either segmental or stresscontrasts, and an adaptation of the segmental-basedspeeded classification task used by Pallier et al. (1997),in which they were required to ignore the irrelevantstress variations.

Concerning sequence repetition, we predicted that, asis typical for French natives, musicians and nonmusi-cians would perform similarly for the segmental con-trast and better for the segmental than for the stresscontrasts. More interestingly, if musicians were better atprocessing the acoustic correlates of stress, their seg-mental over stress contrast advantage should bereduced in comparison to nonmusicians, especially fora weak stress contrast that does not involve strong vocal-ic reduction. In the speeded classification task, we pre-dicted that the interference entailed by the irrelevantand unpredictable stress variations on the segmental-based response would be stronger for musicians thanfor nonmusicians.

Method

Participants

French native speakers without hearing problems partic-ipated in the experiment in exchange for course credits(24) or payment (37). Data for six participants were dis-carded: two because they were bilinguals, two nonmusi-cians because they did not reach the success criterion inthe sequence repetition task (see Procedure), and twobecause their music experience exceeded our selectioncriteria for nonmusicians, which were (i) no solfeggiolessons (ii) never learned to read and write music, andwere unable to do so (iii) no more than three years ofmusic practice, stopped for at least six years. The 32remaining nonmusicians (9 males and 23 females, aged 18to 27 yrs, average: 20 yrs) were tested at the UniversitéLibre de Bruxelles. The 25 musicians (12 males and 13females, aged 21 to 30 yrs, average = 25 yrs) were select-ed at the Conservatoire Royal de Bruxelles. They hadbegun music lessons between 4 and 12 yrs of age (on


2In Dutch, stress contrasts are usually accompanied by contrasts insyllable weight, whereby stress is assigned to the heavier syllable (e.g.,Van der Hulst, 1984).

Music2603_07 1/16/09 9:48 AM Page 237

average, at 7 yrs) and had between 12 and 22 yrs of musictraining (16 yrs, on average). Most were educated in clas-sical music, the majority (16) playing violin, the othersplaying the piano or other instruments. Most (16) wereabsolute pitch possessors (AP). Since most French speak-ers in Belgium are exposed to a stressed language(Dutch), proficiency in Dutch was assessed in all partici-pants. According to their naming scores of 48 standard-ized drawings (Snodgrass & Vanderwart, 1980), bothgroups displayed poor knowledge of Dutch, with 17 and20% correct for nonmusicians and musicians, respec-tively, F < 1, as well as poor knowledge of English, with28 and 33% correct, F(1, 55) = 2.56, p > .10.

Materials

SEQUENCE REPETITION TASK

Three minimal pairs of nonwords were constructed,one involving a segmental contrast (IMbect–OMbect,’imbεkt/–/’ mbεkt/), another a strong stress contrast,

namely a contrast with strong vocalic reduction(Odrept– oDREPT, /’ drεpt/–/ ’drεpt/), and the third aweak stress contrast, namely a contrast with weak vocalicreduction (DRUSnee–drusNEE, /’dr sni/–/dr ’sni:/). Thenonwords were derived from English words: the imbectpair was derived from invent, and the stress contrast pairsfrom OBject–obJECT and TRUSty–trusTEE. A femalenative speaker of Canadian English recorded each non-word several times in a sound-treated booth. Tentokens of each nonword were used so as to promoteacoustic and phonetic variability. As illustrated inFigure 1, in the stress contrast pairs, stressed vowelswere longer (F(1, 36) = 725.45, p < .0001) and louder(F(1, 36) = 231.69, p < .0001) than unstressed ones andhad higher pitch, F(1, 36) = 382.42, p < .0001. Stressedand unstressed vowels differed more in the strongthan in the weak stress contrast pairs in terms ofintensity (F(1, 36) = 4.12, p < .05) and length (F(1, 36) =16.25, p < .0005), as well as by their vocalic quality(Figure 1d).


140

150

160

170

180

190

200

210

IMbect OMbect DRUSnee

Nonword

NonwordNonword

Nonword

(a)

(c)(d)

drusNEE Odrept oDREPT

Pitc

h (H

z)

(b)

60

62

64

66

68

70

72

74

76

IMbect OMbect DRUSnee drusNEE Odrept oDREPT

Inte

nsity

(dB

)

0

50

100

150

200

250

300

350

400

450

IMbect OMbect DRUSnee drusNEE Odrept oDREPT

Syl

labl

e Le

ngth

(m

s)

200

300

400

500

600

700

800

900700 900 1100 1300 1500 1700 1900

F2

F1

FIGURE 1. Average pitch (1a), intensity (1b), syllable length (1c) and formant values (1d) of each syllable of each nonword used in sequence repetition.In Figures a, b, and c, black triangles = syllable 1; transparent squares = syllable 2. In Figure 1d, F1 and F2 values of the stimuli are represented in hertzonly for syllable 1, with the initial stress version of each nonword displayed on the left. Plain lines: “odrept” stimuli; dotted lines: “drusnee” stimuli.

Music2603_07 1/16/09 9:48 AM Page 238

For each minimal pair, five experimental blocks wereconstructed. Each block was made out of 12 trials,except the first block that presented only eight trials. Inthis first block, each trial, namely each sequence thathad to be reproduced, contained two nonwords. In theother blocks, trials consisted of sequences of increasinglength, from 3-nonword sequences in block 2 up to 6-nonword sequences in block 5.

As done in Dupoux et al. (2001), we used a shortsilence interval (30 ms) between successive nonwordsto discourage participants to use recoding strategies,e.g., to mentally translate the nonwords into their asso-ciated key values while listening to the sequence.

SEGMENTAL-BASED SPEEDED CLASSIFICATION TASK

The nonwords mispardof and misparkof were chosen,given that both would make phonologically acceptableitems in English and would thus be easily pronounce-able by the native speaker of Canadian English (thesame as for the previous task) who recorded the mate-rial. This speaker recorded several tokens of these non-words, half with stress on the first syllable (MISpardof,MISparkof, /’mispardof/, /’misparkof/), and half withstress on the second syllable (misPARdof, misPARkof,/mis’pa:rdof/, /mis’pa:rkof/). Ten tokens of each stimu-lus were chosen so as to maximize differences in dura-tion (F(1, 36) = 542.49, p < .0001), intensity (F(1, 36) =76.28, p < .0001), and F0 (F(1, 36) = 837.88, p < .0001)between the stressed and unstressed syllables.

Both materials were digitized at 16 kHz and 16 bits,edited, and stored on a computer.

Procedure

Each participant was tested individually in a quietroom, facing a Macintosh computer with an Azertykeyboard. Presentation of the stimuli and responserecording were controlled using PsyScope (Cohen,MacWhinney, Flatt, & Provost, 1993). Stimuli weredelivered through two loudspeakers located in front ofthe participant, one on the left, the other on the right.

In the sequence repetition task, all participants werefirst presented with the segmental contrast, which wasused to establish baseline performance. Presentationorder of the strong and weak stress contrasts was coun-terbalanced between participants. Participants couldlisten to the various tokens of these items by pressingthe corresponding keyboard keys as many times as theywanted. IMbect was associated to the key “q” andOMbect to the key “ù.” The nonwords with stress on thefirst syllable, Odrept and DRUSnee, were associated to“q.” Those with stress on the second syllable to “ù.” Inthe training phase, participants heard a token of one of

the nonwords and had to press the associated key. Anauditory beep informed them when they made anerror. The success criterion was eight correct responsesin a row for one particular nonword. Those who suc-ceeded turned then to the experimental phase, theirtask being to reproduce each sequence by typing theassociated keys in the correct order. Order of the non-words within sequences excluded long repetitivesequences (e.g., qqq, ùùùù). Order of sequences withineach block was randomized for each participant. Toprevent participants from using echoic memory, oneach trial they had to wait 30 ms after the last nonwordfor a beep and the word “OK” to appear on the com-puter screen before beginning to type their response.During this phase, no feedback was provided. A 2-spause separated the end of the reproduction of thesequence from the next trial.

In the speeded classification task each trial startedwith the auditory presentation of a nonword that par-ticipants had to identify. They indicated their responseby pressing one of two keys corresponding respectivelyto mispardof (q) and misparkof (ù), irrespective of stressposition. The initial consonant of the third syllable wasthe only relevant difference. Belated responses (occur-ring more than 1500 ms after stimulus onset) were sig-naled by auditory feedback. A 200-ms pause separatedresponse from the next trial. There were two variedblocks and two fixed blocks of 64 trials each. In the var-ied blocks, four items (MISpardof and misPARdof vs.MISparkof and misPARkof ) were used and presented ina random sequence. In the fixed blocks, only two itemswere used, keeping the stress pattern constant. Onefixed block contained nonwords with initial stress(MISpardof vs. MISparkof ), the other nonwords withfinal stress (misPARdof vs. misPARkof). Block order wascounterbalanced across participants; item order withina block was randomized for each participant. Thisexperimental phase was preceded by a learning phasefor acquisition of the association of each nonword witha particular key, followed by 16 practice trials.

Participants were tested in a single session, which lastedabout one hour. First, they answered to a questionnaireassessing their music training and expertise. Then, theyperformed the speeded classification task, after which,using a naming task, their knowledge of Dutch wasassessed to verify that they were indeed French mono-linguals. Finally, they performed the sequence repetitiontask, after which their knowledge of English wasassessed in the same way as before for Dutch. Thesequence repetition task was run after the segmental-based classification task in order to reduce the possibil-ity that having to process a stress contrast first wouldmake it more difficult to ignore this speech dimension.


Music2603_07 1/16/09 9:48 AM Page 239

Results

Sequence Repetition

Mean percentages of correct responses (fully correcttranscriptions of the input sequences) are presented inFigure 2, separately for each length, contrast, andgroup. One-tailed t-tests showed that performance wasalways above chance level,3 all ps < .0001. The ANOVArevealed better performance for shorter than longersequences, F(4, 220) = 205.37, p < .0001, and better per-formance for musicians than for nonmusicians, F(1,55) = 25.88, p < .0001. It also showed an effect of con-trast, F(2, 110) = 74.03, p < .0001, with better perform-ance for the segmental (F(1, 110) = 83.82, p < .0001)and strong F(1, 110) = 99.9, p < .0001) stress contrasts(that did not differ from each other, F < 1) than for theweak stress contrast. There were also significant inter-actions between contrast and sequence length, F(8, 440) =3.54, p < .0005, between group and contrast, F(2, 110) =7.34, p = .001, and between group, contrast, and sequencelength, F(8, 440) = 2.19, p < .05. Decomposition of thelatter interaction (using Bonferroni-corrected alpharate of .017) showed that while for the segmental andstrong stress contrast groups tended to differ from each

other only on the longest sequences (interaction withsequence length: F(4, 220) = 3.82, p = .005 and F(4, 220)= 3.01, p = .019), they differed at all sequence lengthsfor the weak stress contrast, F(1, 55) = 19.16, p < .0001(interaction with length: F < 1), with musicians always sig-nificantly outperforming nonmusicians (all ps ≤ .005).4

The fact that groups differed on all long sequences,even for the segmental contrast (from sequence length 4on, all ps ≤ .05) induces a potential confound. In orderto obtain more convincing evidence on the specificeffect that music training might have on the processingof stress contrasts, further analyses were performed on adifference score similar to the one used by Dupoux et al.(2001). This was calculated as performance on the seg-mental contrast minus performance on the stress con-trast. Hence, any remaining length effect observed onthe difference score may be attributed to specific diffi-culties for stress at longer sequence lengths.

Since the weak stress contrast was more difficult thanthe strong stress contrast, two separate difference scoreswere calculated: the segmental advantage was calculated


3According to the binomial law, chance level is equal to 1/2n withn being the sequence length. The chance level is thus 1/4 = 25% atlength 2, 1/8 = 12.5% at length 3, 1/16 = 6.25% at length 4, 1/32 =3.12% at length 5, and 1/64 = 1.56% at length 6.

30

40

50

60

70

80

90

100

2 3 4 5 6 2 3 4 5 6

% C

orre

ct

Sequence Length

Nonmusicians Musicians

Segmental Contrast Strong Stress Contrast Weak Stress Contrast

FIGURE 2. Mean percentages of correct responses in sequence repetition, separately for each length, contrast, and group.

4A separate ANOVA revealed that there was no difference betweenAP possessors and other musicians regarding stress deafness. APpossessors did not outperform other musicians overall, F < 1.Although they tended to perform better than the others on thelongest sequences (on average, 74.7% vs. 62.4% for other musicians),the group by sequence length interaction only tended towards sig-nificance, F(4, 92) = 2.13, p = .08. More crucially, contrast did notinteract with group, and the interaction between group, length, andcontrast was not significant, F < 1 in both cases.

Music2603_07 1/16/09 9:48 AM Page 240

in comparison to either the strong or the weak stresscontrast. Whereas both groups obtained similar per-formance for the segmental and strong stress contrasts,and hence no segmental advantage in this case (Figure 3a),the performance advantage of the segmental over theweak stress contrast, reflecting stress deafness for thelatter, was much less pronounced in musicians. A MannWhitney U test on the difference scores (cf. Peperkamp etal., 1999) showed no overlap between the distributions

of the segmental advantage displayed by musicians andnonmusicians with the weak stress contrast, U = 186.0,p = .0006, this not being the case with the strong stresscontrast, U = 330.5, p = .263.

An ANOVA performed on the difference scores, cal-culated separately for each sequence length, showedsignificant effects of group, F(1, 55) = 11.71 p < .005,and length, F(4, 220) = 3.78, p < .01, as well as a sig-nificant difference between the two difference scores,


5

15

25

35

45

Diff

eren

ce S

core

s (in

%)

Sequence LengthStrong Stress Contrast

2 3 4 5 6 2 3 4 5 6

Weak Stress Contrast

5

−5

−15

−15

−5

15

25

35

45

Diff

eren

ce S

core

s (in

%)

Strong Stress Contrast

(a)

(b)

Weak Stress Contrast

Nonmusicians Musicians

FIGURE 3. Difference scores in sequence repetition for the segmental vs. strong stress contrast and for the segmental vs. weak stress contrastcomparisons, calculated either (a) overall (error-bars represent the standard deviation) or (b) separately for each sequence length.

Music2603_07 1/16/09 9:48 AM Page 241

F(1, 55) = 67.53, p < .0001. There were significant inter-actions between length and type of difference score,F(4, 220) = 4.28, p < .005, as well as between group andtype of difference score, F(1, 55) = 5.80, p < .025, butneither the interaction between group and sequencelength, F(4, 220) = 1.91, p >.10, nor the three factorsinteraction, F < 1, reached significance. Decompositionof the group by type of difference score interaction(using Bonferroni-corrected alpha rate of .025) showedthat group did not differ for the comparison betweenthe segmental and strong stress contrasts, F(1, 55) =2.54, p >.10, whereas a highly significant group differ-ence was observed for the comparison between the seg-mental and weak stress contrasts, F(1, 55) = 9.67, p <.005, without interaction with length, F < 1. Indeed,when considering the segmental and weak stress con-trasts, stress deafness was weaker in musicians at allsequence lengths (Figure 3b).5

Speeded Classification

The ANOVA on RTs to correct responses showed signifi-cant effects of group, F(1, 55) = 11.49, p < .005, and blocktype,6 F(1, 55) = 14.57, p < .005, as well as an interactionbetween these factors, F(1, 55) = 4.99, p < .05. As illus-trated in Figure 4, only musicians suffered from interfer-ence, F(1, 24) = 33.46, p < .0001 (nonmusicians: F = 1).

However, interpretation of this effect is made difficultby the fact that musicians were faster than nonmusiciansoverall, on both varied (F(1, 55) = 5.54, p < .025) andfixed (F(1, 55) = 15.59, p < .0005) blocks.

On accuracy, musicians also outperformed nonmusi-cians, with 98.9 vs. 97.5% correct, respectively, F(1, 55)= 5.82, p < .025. Performance was better overall forfixed than varied blocks, with 98.5 vs. 97.8% correct,respectively, F(1, 55) = 12.34, p < .001, but this effectdid not depend on group, F < 1.7


850

900

950

1000

1050

1100

1150

Musicians Nonmusicians

RT

s (m

s)

Fixed Blocks Varied Blocks

FIGURE 4. RTs to correct responses in speeded classification, separately for fixed and varied blocks in each group (error bars represent the standarddeviation).

6Preliminary analysis showed no difference between the two kindsof fixed blocks (stressed on the first vs. second syllable: 1033 vs. 1035 mson average, respectively).

7Again, no difference was found between AP possessors and othermusicians. AP possessors did not outperform and were not morerapid than other musicians, and group did not interact with blocktype in the analysis on either RTs or correct scores, all ps > .10.

5In agreement with the previous analysis, no difference was foundbetween AP possessors and other musicians, be it in the distributionof the overall segmental advantage (both Mann-Whitney U testswere nonsignificant, p >.10) or in the ANOVA considering sequencelength as an additional variable (F < 1 for the group effect as well asfor the interactions involving group).

Music2603_07 1/16/09 9:48 AM Page 242

Since musicians outperformed nonmusicians overall interms of both response speed and accuracy, interferencemay have been attenuated in the latter because theywere responding more slowly and more poorly even inthe fixed blocks. Amount of interference was actuallyrelated to overall response speed: the correlationbetween the size of the interference on RTs and the RTsobserved on the fixed blocks was significant, r (55) =−.46, p < .0005, at least in nonmusicians, r (30) = −.46,p < .01 (in musicians: r(23) = −.14, p > .10). We there-fore tried to match the two groups on their fixed blockscores.8 In doing this, 20 musicians and 18 nonmusi-cians were found to obtain similar average RTs in thefixed blocks (1018 and 1013 ms, respectively). TheANOVA on these scores showed only an effect of con-dition, with shorter RTs in the fixed than in the variedblocks (1016 vs. 1042 ms, respectively, F(1, 36) = 18.71,p < .0001), without interaction with group, F < 1.Interference was about the same in musicians (28 ms)and nonmusicians (25 ms).

Discussion

The results of the sequence repetition task show that, innatives of a nonstress language, music expertise enhancessensitivity to stress contrasts that do not present majorsegmental correlates. Indeed, musicians exhibitedreduced stress deafness for the weak stress contrast,whereas no group difference was observed for thestrong stress contrast, in which vocalic reductioninduced major segmental correlates to stress, and whichpresented stronger syllable length and intensity differ-ences as a function of stress. This is a new finding sup-porting the idea that processing of difficult linguisticprosodic variations benefits from music training.

Reduced stress deafness in musicians was observed atall sequence lengths, suggesting that nonmusicians’ diffi-culties with stress arise at the encoding of these contrastsrather than from memory. Consistently, nonmusicans’difficulties with stress were already observed in the phaseof learning the association of the isolated nonwords withresponse keys: they were slower to reach the criterion ofeight correct responses in a row (nine participants neededmore than 20 training trials, among whom seven neededmore than 30 trials, which was never the case for musi-cians, who always needed less than 20 trials to distinguishthe stressed words) and their performance was poorer:91%, while musicians reached 98% correct. Dupoux et al.

(2008) also reported an effect at training: their Frenchparticipants showed an effect of contrast, due to moreerrors for stress than for the segment. The late learners ofSpanish showed a smaller but significant effect, and theSpanish no effect at all. Thus, in both studies the groupdifferences do not depend on memory load and are quitestrong.

Still, the results of the speeded classification taskshow that the musicians’ enhanced sensitivity to stresscontrasts does not lead them to be poorer than non-musicians when stress variations must be ignored. Infact, only musicians presented interference from irrele-vant stress variations in this task, but it could be due totheir overall higher response speed. Indeed, whengroups were matched on the fixed block scores, andhence on overall performance level, no difference in theamount of interference was observed anymore.

Taken together, the results of the two experimentssuggest that music training entails a benefit withoutcost pattern in the processing of prosodic features likelexical stress. Although far more sensitive than non-musicians to lexical stress variations, as shown by theirreduced stress deafness phenomenon, musicians areable to discard such variations when they are irrelevantto the task, as it was the case in speeded classification.

Positive associations between music training andspeech processing are sometimes referred to as domain-transfer effects that contradict a modular view ofspeech and music processing. We believe, however, thatthe debate is still open regarding the origin of sucheffects. First, when comparing musicians and nonmusi-cians, observed associations could be either genetic orthe consequence of music lessons. As a matter of fact,musicianship is likely to be a consequence of bothmusic aptitude and training combined with other fac-tors. Second, many data indicate that music and lan-guage have distinct cortical representations and thateither domain can be selectively impaired (see review ine.g., Peretz, 2006). Even the ability to process prosodicinformation such as linguistic intonation seems at leastpartly independent from music processing. Indeed,patients who are impaired in music processing as a con-sequence of a brain lesion and individuals who sufferfrom lifelong difficulties with music do not necessarilypresent severe problems in processing linguistic intona-tions (e.g., Ayotte, Peretz, & Hyde, 2002; Patel, Wong,Foxton, Lochy, & Peretz, 2008; Peretz et al., 1994).While it is possible that only some aspects of languageand music processing are modular (e.g., Peretz &Coltheart, 2003), and that prosody does not belong tosuch modular subsystems, additional evidence is neededon this point. Third, as argued by Schellenberg and


8We discarded the nonmusicians with fixed block RTs higher thanthe maximum displayed by musicians, as well as the musicians withfixed block RTs lower than the minimum observed in nonmusicians.

Music2603_07 1/16/09 9:48 AM Page 243

Peretz (2008), observed associations between musicand language could provide evidence against the mod-ular position, but could also be the product of domain-general attentional or corticofugal influences.

For example, taking music lessons could be onelearning experience that improves executive functionand, consequently, test-taking abilities in a variety ofcognitive domains. Some results of the present studysupport this view. Indeed, in the sequence repetitiontask musicians were better than nonmusicians on alllong sequences, even for the segmental contrast. Thisconverges with former evidence of a positive associa-tion between music training and memory performance(e.g., Chan, Ho, & Cheung, 1998; Ho, Cheung, & Chan,2003; Kilgour, Jakobson, & Cuddy, 2000). In addition,since musicians were far more sensitive than nonmusi-cians to lexical stress variations, as shown by theirreduced stress deafness phenomenon, one would haveexpected them to be less able to discard such variationswhen they are irrelevant to the task. This was not whatwe observed in speeded classification. This task actuallyrequired selective attention to the target dimension,here a segmental one. Thus, the fact that musicianswere as able as nonmusicians to discard irrelevant stressvariations (at least when they performed at the samelevel overall) suggests that musicianship does not onlytune encoding of some linguistically relevant parame-ters, but also develops, in parallel, the capacity to attendselectively to auditory dimensions, whatever thedomain. Analytic skills developed through music train-ing could be at the core of this effect.

Whether general cognitive skills can also explain theincreased sensitivity to lexical stress contrast exhibitedby musicians in the sequence repetition task should befurther investigated. As already commented on, thiseffect does not seem to depend on improved memorycapacity. In fact, both the observation of the partici-pants’ behavior during the sequence repetition task anddata from debriefing suggest that musicians used somemusic strategies to perform the task. Indeed, when firstpresented with the stress contrasts, several musiciansspontaneously referred to intensity and/or durationdifferences between the stressed and unstressed sylla-bles. Although some nonmusicians also noticed suchdifferences, only musicians frequently referred to the“rhythm” of the sequences, and during the task, some ofthem presented a rhythmic balancing or noddingbehavior never adopted by nonmusicians. However,such music strategies may depend on a general execu-tive function improvement, allowing musicians, forexample, to better focus their attention on duration orintensity variations between the stressed and unstressed

syllables of the weak stress contrast. In other words,although in the present study the stress deafness effectdisplayed by musicians was not significantly correlatedwith the amount of interference they presented inspeeded classification (on RTs: r = −.21; on correctscores: r = −.11, p >.10 in both cases), both effects mayin principle depend on a general executive functionimprovement induced by musicianship.

In any case, the fact that musicianship affords anadvantage for lexical stress processing may have impor-tant educational consequences. For example, we knowthat learning a new language, especially in the initialphase wherein one needs to segment new words, maylargely benefit from the motivational and structuringproperties of music in songs (Schön et al., 2008). Futurework should examine whether the effect observed hereis an aid for the acquisition of a foreign language.

It has been reported that music ability facilitates theacquisition of a second language sound structure byimproving receptive and productive phonological abil-ities (Slevc & Miyake, 2006). Late arrival Japaneseadults (living in the United States) who were good atanalyzing, discriminating, and remembering musicstimuli also were better than their less musically talentedpeers at accurately perceiving and producing sounds oftheir second language (English). Yet, Slevc and Miyakeonly looked at the ability to discriminate segmentalcontrasts, by asking participants to decide which mem-ber of a minimal pair like “playing/praying” was pre-sented in contextually neutral sentences.

What the present results suggest is that music train-ing may favor second language learning by improvingthe ability not only to process non-native segmentalcontrasts (as shown by Slevc & Miyake, 2006) but alsoto process (or to pay attention to) non-native prosodiccontrasts. Our finding that musicians show increasedsensitivity to stress contrasts in a foreign language (theEnglish-like nonwords they were presented with) isconsistent with the fact that musicians are more sensi-tive than nonmusicians to pitch variations in intona-tion contours for a foreign language that they do notunderstand (Marques, Moreno, Castro, & Besson, 2007).

One way to assess whether such effects of music train-ing afford substantial benefit to second language acquisi-tion may be to compare French musicians andnonmusicians in a word-learning paradigm to study howeasily they can acquire new lexical items that differ eitherin only their stress patterns, or in only their pitch patterns.Indeed, presenting native English speakers with Englishpseudosyllables superimposed with three non-nativesuprasegmental contrasts (pitch patterns), Wong andPeracchione (2007) observed large individual differences


Music2603_07 1/16/09 9:48 AM Page 244

in learning success that were associated with the learn-ers’ ability to perceive pitch patterns in a nonlexical con-text and their previous music experience. We wouldpredict a similar strong association as regards both lexi-cal stress and pitch patterns. In addition, the positiveeffect of music training could generalize to reading abil-ities, since stress processing abilities influence readingdevelopment in a second, stress-based language (Goetry,Wade-Woolley, Kolinsky, & Mousty, 2006).

Author Note

Preparation of this article was supported by a grantfrom the Human Frontier of Science Program (RGP53/2002, “An interdisciplinary approach to the problem

of language and music specificity”) and by a FRFCgrant (2.4633.66, “Mental representations of music andlanguage in singing and nature of their interactions”).The first author is Senior Research Associate of theFonds de la Recherche Scientifique—FNRS.

Many thanks to Leslie Wade-Woolley for having pro-duced the material, to Pascale Lidji and two anonymousreviewers for their helpful comments on a previous ver-sion of this paper, and to Emmanuel Bigand for hisattentive and friendly organization.

Correspondence concerning this article should beaddressed to Régine Kolinsky, UNESCOG Universitélibre de Bruxelles, CP 191Av. Franklin Roosevelt, 50B-1050 Brussels, Belgium. E-MAIL: [email protected]


References

ABRAHAM, G. (1974). The tradition of Western music (pp.

62-83). Berkeley, CA: University of California Press.

AYOTTE, J., PERETZ, I., & HYDE, K. (2002). Congenital amusia:

A group study of adults afflicted with a music-specific disorder.

Brain, 125, 238-251.

BECKMAN, M. E., & PIERREHUMBERT, J. (1986). Intonational

structure in Japanese and English. Phonology Yearbook, 3, 255-

309.

BIGAND, E., TILLMANN, B., POULIN, B., D’ADAMO, D. A., &

MADURELL, F. (2001). The effect of harmonic context on

phoneme monitoring in vocal music. Cognition, 81, B11-B20.

CHAN, A. S., HO, Y.-C., & CHEUNG, M.-C. (1998). Music

training improves verbal memory. Nature, 396, 128.

COHEN, J. D., MACWHINNEY, B., FLATT, M., & PROVOST, J.

(1993). PsyScope: An interactive graphic system for designing

and controlling experiments in the psychology laboratory

using Macintosh computers. Behavior Research Methods,

Instruments, and Computers, 25, 257-271.

DANCOVICOVÁ, J., HOUSE, J., CROOKS, A., & JONES, K. (2007).

The relationship between musical skills, music training,

and intonation analysis skills. Language and Speech, 50,

177-225.

DUPOUX, E., PALLIER, C., SEBASTIÁN, N., & MEHLER, J. (1997).

A destressing ‘deafness’ in French? Journal of Memory and

Language, 36, 406-421.

DUPOUX, E., PEPERKAMP, S., & SEBASTIÁN-GALLES, N. (2001).

A robust method to study stress “deafness.” Journal of the

Acoustical Society of America, 110, 1606-1618.

DUPOUX, E., SEBASTIÁN-GALLÉS, N., NAVARRETE, E., &

PEPERKAMP, S. (2008). Persistent stress ‘deafness’: The case

of French learners of Spanish. Cognition, 106, 682-706.

GARDE, P. (1968). L’accent [The accent]. Paris: Presses

Universitaires de France.

GARNER, W. R. (1974). The processing of information and

structure. Potomac: Erlbaum.

GOETRY, V., WADE-WOOLLEY, L., KOLINSKY, R., & MOUSTY,

P. (2006). The role of stress processing abilities in the

development of bilingual reading. Journal of Research in

Reading, 29, 349-362.

GRABE, E., & LOW, E. L. (2002). Durational variability in

speech and the rhythm class hypothesis. In C. Gussenhoven

& N. Warner (Eds.), Laboratory phonology 7 (pp. 515-546).

Berlin: Mouton de Gruyter.

HO, Y.-C., CHEUNG, M.-C., & CHAN, A. S. (2003). Music

training improves verbal but not visual memory: Cross-

sectional and longitudinal explorations in children.

Neuropsychology, 17, 439-450.

HURON, D., & OLLEN, J. (2003). Agogic contrast in French

and English themes: Further support for Patel and Daniele

(2003). Music Perception, 21, 267-271.

JONES, M. R., & BOLTZ, M. (1989). Dynamic attending and

responses to time. Psychological Review, 96, 459-491.

KILGOUR, A. R., JAKOBSON, L. S., & CUDDY, L. L. (2000).

Music training and rate of presentation as mediators of text

and song recall. Memory and Cognition, 28, 700-710.

KLATT, D. (1976). Linguistic uses of segmental duration in

English: Acoustic and perceptual evidence. Journal of the

Acoustical Society of America, 59, 1208-1221.

LEHISTE, I. (1970). Suprasegmentals. Cambridge, MA: MIT

Press.

LEHISTE, I., OLIVE, J. P., & STREETER, L. (1976). The role of

duration in disambiguating syntactically ambiguous

sentences. Journal of the Acoustical Society of America, 60,

1199-1202.

LERDAHL, F. (2001). The sounds of poetry viewed as music.

Annals of the New York Academy of Sciences, 930, 337-354.

Music2603_07 1/16/09 9:48 AM Page 245

LOW, E. L., GRABE, E., & NOLAN, F. (2000). Quantitative

characterizations of speech rhythm: Syllable-timing in

Singapore English. Language and Speech, 43, 377-401.

MAGNE, C., SCHÖN, D., & BESSON, M. (2006). Musician

children detect pitch violations in both music and language

better than nonmusician children: Behavioral and electro-

physiological approaches. Journal of Cognitive Neuroscience,

18, 199-211.

MARQUES, C., MORENO, S., CASTRO, S.-L., & BESSON, M.

(2007). Musicians detect pitch violation in a foreign language

better than nonmusicians: Behavioral and electrophysiological

evidence. Journal of Cognitive Neuroscience, 19, 1453-1463.

NARMOUR, E. (1990). The analysis and cognition of basic melodic

structures. Chicago, IL: University of Chicago Press.

NILSONNE, A., & SUNDBERG, J. (1985). Differences in ability of

musicians and nonmusicians to judge emotional state from

the fundamental frequency of voice samples. Music Perception,

2, 507-516.

PALLIER, C., CUTLER, A., & SEBASTIÁN-GALLÉS, N. (1997).

Prosodic structure and phonetic processing: A cross-

linguistic study. Proceedings of Eurospeech ’97, 5th European

Conference on Speech Communication and Technology, 4,

2131-2134.

PALMER, C. (1996). On the assignment of structure in music

performance. Music Perception, 14, 23-56.

PALMER, C. (1997). Music performance. Annual Review of

Psychology, 48, 115–138.

PALMER, C., JUNGERS, M. K., & JUSCZYK, P. W. (2001).

Episodic memory for musical prosody. Journal of Memory

and Language, 45, 526–545.

PALMER, C., & KELLY, M. H. (1992). Linguistic prosody and

musical meter in song. Journal of Memory and Language, 31,

525-542.

PATEL, A. D., & DANIELE, J. R. (2003a). An empirical

comparison of rhythm in language and music. Cognition,

87, B35-B45.

PATEL, A. D., & DANIELE, J. R. (2003b). Stress-timed vs.

syllable-timed music? A comment on Huron and Ollen

(2003). Music Perception, 21, 273-276.

PATEL, A., WONG, M., FOXTON, J., LOCHY, A., & PERETZ, I.

(2008). Speech intonation perception deficits in musical tone

deafness (congenital amusia). Music Perception, 25, 357-368.

PEPERKAMP, S., DUPOUX, E., & SEBASTIÁN-GALLÉS, N.

(1999). Perception of stress by French, Spanish, and bilingual

subjects. Proceedings of Eurospeech ’99, 6, 2683-2686.

PERETZ, I. (2006). The nature of music from a biological

perspective. Cognition, 100, 1-32.

PERETZ, I., & COLTHEART, M. (2003). Modularity of music

processing. Nature Neuroscience, 6, 688-691.

PERETZ, I., KOLINSKY, R., LABREQUE, R., TRAMO, M.,

HUBLET, C., DEMEURISSE, G., & BELLEVILLE, S. (1994).

Functional dissociations following bilateral lesions of

auditory cortex. Brain, 117, 1283-1301.

PRICE, P. J., OSTENDORF, M., SHATTUCKHUFNAGEL, S., &

FONG, C. (1991). The use of prosody in syntactic disambiguat-

ion. Journal of the Acoustical Society of America, 90, 2956-2970.

RAMUS, F., NESPOR, M., & MEHLER, J. (1999). Correlates of

linguistic rhythm in the speech signal. Cognition, 73, 265-292.

SCHELLENBERG, E. G., & PERETZ, I. (2008). Music, language

and cognition: unresolved issues. Trends in Cognitive Sciences,

12, 45-46.

SCHÖN, D., BOYER, M., MORENO, S., BESSON, M., PERETZ, I.,

& KOLINSKY, R. (2008). Songs as an aid for language acquisi-

tion. Cognition, 106, 975-983.

SCHÖN, D., MAGNE, C., & BESSON, M. (2004). The music of

speech: Music training facilitates pitch processing in both

music and language. Psychophysiology, 41, 341-349.

SCOTT, D. R. (1982). Duration as a cue to the perception of a

phrase boundary. Journal of the Acoustical Society of America,

71, 996-1007.

SLEVC, L. R., & MIYAKE, A. (2006). Individual differences in

second language proficiency: Does musical ability matter?

Psychological Science, 17, 675-681.

SLOBODA, J. A. (1985). Expressive skill in two pianists: Metrical

communication in real and simulated performances. Canadian

Journal of Psychology, 39, 273-293.

SNODGRASS, J. G., & VANDERWART, M. A. (1980). A

standardized set of 260 pictures: Norms for name agreement,

familiarity and visual complexity. Journal of Experimental

Psychology: General, 6, 174-215.

THOMPSON, W. F., SCHELLENBERG, E. G., & HUSAIN, G.

(2004). Decoding speech prosody: Do music lessons help?

Emotion, 4, 46-64.

TODD, N. (1985). A model of expressive timing in tonal music.

Music Perception, 3, 33-58.

TRIMMER, C. G., & CUDDY, L. L. (2008). Emotional intelli-

gence, not music training, predicts recognition. Emotion, 8,

838-849.

VAN DER HULST, H. (1984). Syllable structure and stress in

Dutch. Dordrecht: Foris.

WENK, B. J. (1987). Just in time: On speech rhythms in music.

Linguistics, 25, 969-981.

WONG, P. C. M., & PERACCHIONE, T. K. (2007). Learning

pitch patterns in lexical identification by native English-

speaking adults. Applied Psycholinguistics, 28, 565-585.

WONG, P., SKOE, E., RUSSO, N., DEES, T., & KRAUS, N.

(2007). Musical experience shapes human brainstem encoding

of linguistic pitch patterns. Nature Neuroscience, 10, 420-422.


Music2603_07 1/16/09 9:48 AM Page 246

music training facilitates lexical stress processing

Documents