MororCohtol, 2005,9,23,39 @20os HumanKihericsPlblhhe6,Inc.

Finger-Tapping Ability in Maleand FemalePianists and Nonmusician Controls TomokoAoki, ShinichiFuruya,and Hiroshi Kinoshita Using fast tapping tasks with each ofthe four fingen (single-finger tapping) and with two of the frngers used altemately (double-finger tapping), the ability to make rapid tapping movement by the individual fingers was compared between expert pianistsand nonmusiciancontrols in both genders.Maximal pinch and grasp forces were also measuredto assessstrength of individual fingers and whole hand, respectively. Movement of the ring and little fingers was slower than that of the index and middle fingers in both the pianists and controls. The slownessof the ring and little fingers was, however,much lessevident in the pianists than the controls in both tapping tasks. The pianists also had smaller intertap interval variability for the index and middle fingers. No pianist--{onftol differcnce was found for the pinch and grasp forces. Piano training, thercfore, effectively changed the ability to move individual fingers rapidly, but not thelr flexor strength.No gender difference was found in any of the tapping tasks though males had greater strength. Gender thus does not appear to be a factor differentiating the ability to move individual fingers rapidly. Key Wordst ttotnirg, piano, gendet difference, nngef tapping Everyday experience tells us that individual fingers differ in their facility ofusage. Clearly, the index (I) and middle (M) fingers are used more dominantly for a wide variety of daily manipulative tasks than the ring (R) and little (L) fingers. Raj and Marquis (1999), who investigated finger dominance by recording the finger used when inputting numbers into a calculator, reported that more than 80% of the participants utilized the I finger, and progressively less usage was observed for the M finger, thumb, L finger, and R finger. In grasping familiar objects, Kamakura, Matsuo, Ishii, Mitsuboshi, and Miura (1980) also reported that the use of either the I or M finger was more commonplace than th€ R and L fingers. Aoki, Francis, and Kinoshita (2003) recently examined the motor ability of individual fingers during fast tapping by the I, M, R, or L finger (single-finger

Theauthors arewith th€craduateSchoolof HumanSciences, OsakaUniversity, t -2Yamadaoka, Suita, Osaka565-0871, Japan.Kinoshitais alsowith theuniversity's Schoolof HealthandSponSciences.

24

Aoki,Furuya,and Knoshita

tapping) and fast altemate tapping by a pair of these fingers (double-fingertapping). The participantsusedin their study were all young maleswho did not have ip"iific finget -ou"ment training suchasplaying the piano or guitar.The resultsof single-fingertappingindicatedthatthe I finger attainedthe highesttappingcadence, foliowed by ttre M, L, and R fingers in descendingorder The magnitudeof the key-striking force was also greatestfor the I finger, followed by the M' L, and R fingers.The double-fingertapping test further indicated that the tapping cadence of eachfinger could changedependingon the paired finger selected.The cadence ofthe IM finger pair (IM combination)was highest,and that oi the RL finger pair (RL combination) was the lowest. Perceivedtask difficulty was also highest for the RL combination. In another recent study using positron-emissiontomography (PET)' Aoki et al. (2004) demonstratedthat activationin the cortical and subcorticalar€aswas much more extensiveduring double-finger tapPingby the RL combination compared with the IM combination or single-fingertapping by the I or R finger, confirming the relatively high central loading for tapping by the RL combination.There are also somerq)orts concerningthe differencein strengthofindividual fingersamong ordinary people.Kinoshita, Murase,and Bandou ( 1996);Radwin, Oh, Jensen,and Webster( 1992); and Zatsio$ky, Li, and Latash ( 1998,2000) measuredthe maxtmum isometric flexion force of individual fingers. They all reported geater strength for the I or M finger than the R or L finger. The difference in the frequency of everyday usage undoubtedly leads to the difference in the level of training, possibly causing an interdigit differencein motor function. This possibility has,however,remainedunexplored.One solution to this problem might be to examinemotor function of the fingers in individuals with a backgroundof extensivemovementtraining of all fingers, and to compare it with that of individuals with no such background.Expert piamsts representa group ofindividuals with many yearsofspecific faining for isolatedmovementof individual fingers,and thus the motor function oftheir fingerscould be optimized' The primary purposeof the presentstudy was to investigatethe ability of expert pianiits to perform fast singl;-finger tapping using the I, M, R, or L finger and fast double-fingertapping by the IM, MR, or RL combination, and to comparethese with thoseof nonmuJiciin controls.Temporaland force parametersdescribingthe movementof tapping finger(s), as well as the intraindividual variability of these parameters,wer;;valuated to assessthe ability to move individual fingers repetiiively and rapidly. The isometric pinch force between the thumb and eachfinger, and the graspfoice with the whole hand were also measuredto determinethe strength ofindividual fingersin the pianists.Long periodsofcontinuous practiceofstriking keys,as well as brief bouts of forceful key-stdking moYementsin the pianistscan lead to improved force production capacity of individual fingers. Another issuethat might needto be re-examinedin relation to motor function of individual fingers is the gendereffect. Studiesindicate that males generally have larger and strongermusclesthan females,and the differencestend to be more pronouncedin the musclesof the upper limbs (Heyward, Johannes-Ellis,& Romer, 1986).Fast type IIA fibers were also found to occupy a greaterareain the musclesin males, whlreas slow type I fibers occupied a greaterarea in females

FingerTapping in Pianists

(Staronet a1.,2000).Therefore,males could be able to use theseto advantagein moving their fingers faster than females.On the other hand, in a recent study of the effectsof force crosstalkamongthe fingersduring isometdc force production tasks,Shinohara,Li, Kang, Zatsiorsky,andLatasb(2003) reportedthat sucheffects were significantly less in femalesthan males, suggestinga female advantagein isolatedfinger movement.Indeed,the findings of previous studieson genderdifferencein finger tapping ability are not consistent.While Nicholson and Kimura (1996),PetersandCampagnaro(1996),and Ruff andParker(1993) all reportedthat malestappedfasterthan females,Hermsdorfer,Marquardt,Wack.and Mai (1999) and Mckeeverand Abramson (1991) found no genderdifferencein a simila.rtask. Nicholson and Kimura also reportedno genderdifferencein a task ofsequentially pressingtelegraphkeys with four fingers.In testingsix different fine motor tasks, Petersand Campagnarofound that the magnitudeand direction of gendereffects on the task performancevaried dependingon the task. The secondarypurposeof the presentstudy was,therefore,to determineifgender influencesthe performance outcomein single-fingertapping by the I, M, R, or L finger and double-fingertapping by the IM, MR, or RL combination.

,Vtethods ParticiDants The participantsin this study were 10 experiencedclassicalpiano players(5 men and 5 women, mean age t SD = 2l.l I 2.3 years) and 10 nonmusiciancontrols (5 men and 5 women, mean age r SD = 20.6 a 3.9 years).All pianistshad played the piano for more than 13 years.All participantswere right-handed,asdetermined by the EdinburghMRC HandednessInventory(Oldfield, 1971),and had no history of neuropathiesor traumato the upper extremities.Written informed consentwas obtained from all participantsbefore the study. The human ethics commtttee at OsakaUniversity approvedthe study,

Measurement of Tapping Performance Four miniature strain-gaugeforce transducerswere used to monitor the striking forces generatedby the tips of the I, M, R, and L fingers during tapping. These tansducers were firmly attachedto an acrylic plate that was bolted to the upper surface of a 68.5 cm high testing table. The distancesseparatingthe centersof adjacenttransducerswer€ 3 cm betweenthe I and M fineers.2.5 cm betwcennc M and R fingers,and 3.5 cm betweenthe R and L fingerslFor rhe thumb, a heightadjustableplasticplate wasplacedadjacentto the I finger tansducer.The apparatus also included a height-adjustableand position-adjustablecircular palm rest,and a height-adjustablefoream rest with restrainingstraps.The height ofthe thumb and forearmrests,and the position ofthe palm restwere adjustedto fit the right handof eachindividual participant.Details ofthe force-measurement apparatushavebeen describedin our previouspapers(Aoki er at., 2003; Aoki & Kinoshrra,2001). After receiving a detailedexplanationof the experimentaltasks,the participant was seatedcomfortably in a chair facing the testing table. The participant's

Aoki,Furuya, and Kinoshita right upper arm was parallel with his or her torso, and the forearm was extended horizontally andparallelto the sagittalplane.Eachparticipantperformedtwo types oftapping tasks:single-fingeranddouble-fingertapping.Single-fingertappingwas performed by each of the I, M, R, and L fingers. Panicipantsused a designated finger to repetitively sftike its coresponding key. Double-fingertapping involved movementswith the I and M, M and R, and R and L fingers.In the double-finger tapping, two designatedfingers were used to altemately strike each of their correspondingkeys. During both tasks,participantswere instructedto tap the key(s) repeatedlywith the finger(s)asrapidly and asconsistentlyaspossible,andto move the finger(s) by emphasizingflexion and €xtension at the metacarpophalangeal Participantswere also requestedto maintain contactbetweenthe tips of the Joint. "resting"(nontapping) fingersand rheir respectivekeys.and betweenthi thumb and its suppon during tapping.The weight ofthe upperlimb was supportedon the forearm rest to minimize its effect on the comDressionforce exertedbv the finsers during tapping.The palm was placedon rhe pilm rest,and movemeniof the w-rist was minirnized by appropriateadjustmentsof the restrainingstrap. To acclimaiethem to the experimentalenvironmentand testingapparatus,as well as to leam to perform eachtappingtask appropriately,participantsengagedin a practice sessiona few days before the experiment.During data collection, each finger or finger-combinationcondition was performedcontinuouslyfor aperiod of 9 s. Only the right hand was testedin this study.The order of conditionswas randomized.Data were collectedduring one trial for eachfinger or finger-combination after severalpracticetdals for each condition. If a participantperceivedthat he or shehad failed to perform as instructed,he or shewas askedto repeatthe tdal. An adequaterest period was given betweentrials- After completing ihe tapping tasks partrcrpantswere requestedto evaluatethe relativ€ "difficulty" of tapping for each of the fingers or finger-combinationsusing an ordinal scalefrom | (easiest)to 4 (mostd.fficult) in the single-fingertapping, or from | (easiest)ut3 (mostdifficult) in the double-fingertapping. Amplified force-transducersignalswere digitized via a 12-bitA/D converter at a sampling rate of 400 Hz for each channel, and then stored on a Macintosh personalcomputer.Beginning 2 s after the initiation oftapping for each trial, data from 10 consecutivetaps by each finger were subjectedto analysis.Five descriptlve parameterswere generatedfrom the data of each tap obtainedfrom each of the tapping finger or fingers(seeFigure l). The first parameterwas "intertap interval": the time duration betweenthe detection of successivefinger-key contacts. Two additional parameterswere generatedby dividing the intertap interval into two components,which were the "contact time"-the time duration of finger-key contact-and the "noncontacttime"-the time durationbetweensuccessivephases of contact. A threshold force of 0.03 N was used to determine the momints of finger key contactandrelease.The fourth parameterwas "averageforce": the force of the tapping finger averagedover the contact period. Averageforce was evaluated insteadof peak force becausethe time-history curve of the tapping force was characterizedby a bimodal pattern in most cases(seeFigure 1), although it was unimodal in somecasesdependingon the trial, finger, and participant.In addition, even in the bimodal curve, the initial peak was not always higher than the second peak.The fifth parameterwas "impulse": the tapping finger force integratedover

-

FingerTapping in Pianists

27

0.4 0.2

(lt) 0.6 0.4

o.2

o

r

( ) 0.6 0.4 6.2

Figure l-Typical time history (single participant) of key-contact forces exerted by tappingfingersduring single-fingerand double.fingertapping.A: Exampleofsingle. finger tapping with the middle finger. B: Example of double-finger tapping with the middle and ring fingers,a: intertap interval, b: contact time, ci noncontacttime, d: averageforce,e: impulse. the contactperiod. For the double-fingertapping tasks,the dataobtainedfrom the two fingers were ayeraged.

Measurementof Pinchand Crasp Forces The maximum isometdc pinch force produced by each tip of the right I, M, R, and L fingers againstthe right thumb for few secondswas measuredusing a selfmade strain-gaugeforce transducer.The thicknessof this transducerwas set at

2a

Aoki,furuya,andKinoshita

15 mm and it weighed 50 g. The force signal from this transducerwas amplified and digitized at 400 Hz and stored in a computer.Data were obtainedfrom two trials for eachfinger for the pinch-forcetask. iJsing an ordinary grasp-forcedynamometer(model TKK-5401, Takei kiki-kogyo Co., Niigata, Japan),the maximum grasp force by the right hand was obtained from each participant for two trials. For eachfinger and hand, the greaterforce value of the two triils was usedfor the subsequentanalys€s.

Statistical Analysis Mean and standarddeyiation values of all parameterswere computed basedon 10 tapsperformedby eachparticipantunder eachof the tappingconditionsexamined. Using th€sevalues,the coefficient of variation (CV) was also computedfor the evaluationofintraindividual variability.A two-way repeated-measurei analysis of variance (ANOVA) was performed to examine the group and gender effects on each parameterfor each finger during single-finger tapping and each finger_ combinationduring double-fingertapping.Two-way repeatedmeasuresANOVAs (groupx gender)were also performedfor the maximum pinch force ofeach finger, and maximum grip force. In addition, using the pooled data of male and female pianists, a one-way repeatedmeasuresANOVA and Tukey's post hoc multiple comparisonswere also performed to examine the finger or finger-combination differencefor each parameter.Statisticalsignificancewas acceptedat p < .05.

Results Single-Finger Tapping Group Difference. The pianists had smaller group mean intertap interval and noncontacttime values for all fingers examined than the nonmusiciancontols (seeFigures2A.and B), indicating that the pianistswere able to move their fingers fasterthar the contols. ANOVAs confirmed that the pianist--controldifferenci in the intertap interval was significant for the M, R, and L fingers but not for the I finger For the noncontacttime, the groupdifferencesfor all fingerswere significant. Comparedwith theseparameters,the pianist-rontrol difference was lessclear in the contact time (Figure 2C). Indeed, none of the fingers had a significant main group effect. Therefore,the fastertapping movementobservedin the pianisB was principally the result of fastermoyementsofthe tappingfinger in space.The group meanvaluesof averageforce, as well asimpulse,were greaterfor the pianisti than for the controls (Figures 2D and E, respectively);however, the di*erence was significantonly for the I and M fingers. Gender Difference. For all temporaland force-parametermeanvaluesofsinglefinger tapping, no genderdifferencewas noted. Finger DilTerencein Pianists. ANOVAs revealeda significantfinger difference in the intertapinterval,contacttime, averageforce, and impulse.posihoc analyses indicatedthat for the intertap interval and contact time, the mean valuesof the R and L fingers were significantly greaterthan those of the I and M finqers. It was

FingerTapping in Pianists

:>:

.E;i I

3;E

! E>

.E -

60 _.: -ll

gEr

o

€.EF"

!

EiE; EiEi c a g

o 8

E

8

R

l

E g!H

( l F

a i

E

> 2 , .

:f€

l

tttf

8

(sut) a(uDlletuoJ

T . Ef i . EH € a: .r F 6 C : E€

E

_

9 , Ct ,Er fl

o

*^:

C

:

.gg: !pi!

t::

!lii

toB

I

I

R

(sur)outt p"luol-uoN

I

ItJ q

E

8

K

i

H, s:

8

.= 9H

(surN) .stndul

. f #he gE EE€ o ; € o

=

i : E$ a

E {

E

j

erE ^E

E*I E I E 3 :

< N R N R E E o 5 S 3 E : I ; (sur) telJotu! deU.lul

(N) erroJ o6Ele^V

EE"E; i r , 9 E

fE:5 , E t iE T f5 s N € =

q ,

E"Fi E

30

Aoki,Futuya,andKinoshita

revealedthat the averagetapping forces by the R and L fingers were significantly lower than those by the I and M fingers. In addition, the impulse of thi R finger was significantly smaller than the I finger

Double-FingerTapping Group Difference. The intertap interval, noncontacttime and contact time for the pianists were less varied acrossall three finger-combinationscomparedwith thoseof the conhol participants,who varied greatly from the IM to RL combinatron (Figures3A, B, andC). The group differencein tappingcadencewastherefore finger-combinationdependent:The differencewassmillest for the IM combination and largestfor the RL combination.ANOVAS revealedthat the group differences in all temporal parameterswere significant for each of the finger-combinations. The meanvaluesof averagetappingforce for the IM and MR combinationstended to be greaterfor the pianiststhan the conhols (Figue 3D), but only the difference for the MR combination was statistically significant.The difference in the mean values of impulse betweenthe pianists and contols was not significant for any of the finger-combinationsalthough the RL combination for the pianists had a relatively small value comparedwith the controls (Figure 3E). Gender Difference. There was no gender difference in any of the temporat and force parametersfor double-fingerlapping. f inger-Combination Difference in Pianists. ANOVAs revealed a sienificant effect of finger-combinationon the intertap irterval, contact time, and impulse. Post hoc analysisindicated that the differencesbetweenall finser combinitions for the intertapinterval and contacttime were significant-It wasalso found that the impulse of the IM combination was significantly smaller than the other two.

I ntraindivid ual VariabiIi ty Gmup Difference. To examinethe pianist--controldifferencein the stabitity of tapprng movements, the mean CV values of intertap interval and averageforce were compared.For single-fingertapping, the mean valuesof intertapinterval for the I and M fingers were significantly lessfor th€ pianiststhan the controls,while those for the R and L fingers did not differ between the two groups (Figure 4A). The CV values of average force were much larger than those for the intertap interval for all fingers (Figure 4B), and none of rhe fingers had a significant group difference in this parameter. For double-finger tapping, neither the intertap interval nor averageforce _ had significant group difference for any of the finger-conbinations (Figures 4C and D)- The CV value of the averageforce in the RL combination was relatively larger for the controls than for the pianists.This differencedid not reachthe level of significance,however Gender Difference. No gender difference was found in any of the CV parameters for both single- and double-fingertapprng.

FingerTapping in Pianists

fl.E

F

E€E E F g i Y.b

.EAs, 54 9: !6

o.o o

oe

8

-

:

E

3

tsE9e E:5;" c H € €

EEEi

++i+

8

(sur) oufr lfeluoJ

U :

:i-: '

EsEt "'l! E E!Ei

o

iE+:

.*te€ 5".F EE I E L 9

S 8 €i E> F g, *= .d; ' €A EE

€iE8 E " E ,bg E . ES g O $ H R H F F

utff

I

E

(sl^|l)ouxl lrPluo)-{oN

b E : F

I

HE€€

(su,N) .stndurl

. l

c o E o o E


E 9 R (sur)te^relu!deuetul

o3

q

\

9

q

{

.

(N) olroJ oOEl.lv

E

E - =

E .aE. E9 Y

(

E

€ EF > EsEa

8 ! ;+

;EgF" F8f-e

E€B ;

E€g€ . E+: E F ! 9 E I E 6DE

1 $ ls"

-

Aoki, furuya, and Kinoshita

EPi

3 ' E v

= €F < E> 1"HA

o

9

t E'i

$EF

FiiE

; ! *

E d G

EFf E* €

++++

o o : :

' E6= 9= : J |E

C

9 .o

o

0

C

o

=

cl8

B

ft

R

9

o s (I) AJ elrol.6ero^v

(X) AO etrioJe6eloav

3>.q

E 9s: *d ' as( Egf !i

3s€q

:9i€

ieE ; d ol a! : J E A E

.e: :; . bE.. :E

Ecs€ r'{ ? i

EE ;$

o

? G ' i . :h U l-e

9 E8 H = l E 9

<:

c)1 (z) AJ te^relu!deuelul

(u ) AJ tE^lelu!dEuelul

r**l EEu.t ie i!SE"n E :-: i" g

#6#E

-

Fin8erTappingin Pianrsts

l3

RankOrder of TaskDifficulty The mean rank orders of task difficulty in single-fingertapping for the I, M, R, andL fingersfor the pianistswere l-2 !0.4 (SD),2.1r 0.6,4.0 r 0, and2.7 r 0.5, respectively.The conesponding valuesfor the conrols were 1.0 t 0, 2.3 t 0.4, 4.0 i 0, and 2.8 r 0.4, respectively.Consequently,the participantsin boto groups equally reported that the R finger was most difficult to move, followed by the L, M. and I fingers in descendingorder. The meanrank ordersof task difficulty in double-fingertapping for the IM, MR, and RL combinationswere 1.0 !0,2.2 !0.3, and2.8 t 0.3 for the pranrsts, respectively,and 1.0 r 0, 2.0 r 0, 3.0 r 0 for the controls,respectively.Therefore, for both groups,the RL combination was most difficult, and the IM combination was the easiest. The rank orders for the single- and double-finger tapping tasks were similar for the men and women.

Maximal Pinchand Crasp Forces Group Difference. The I and M fingers produced greater force than the R and L fingen for both groups (Figure 5). No significant group difference was found in the maximum pinch force betweenthe thumb and any of the individual fingers. The mean valuesofthe maximum graspforce for the pianistswere 280.3* 36.6 N in women and 475.3 r 112.1N in men. The correspondingvaluesfor the controls werc275.2 ! 18.9 N in women and 439.2 r 61.3 N in men. ANOVA revealedno significant differencebetweenthe pianist and control groups. Gender Difference. The pinch forces ofall fingers were significantly greater for malesthan females(Figure 5). Note that the gendereffect was more evidentfor the I and M fingers than the R and L fingers. As indicat€d above, the mean grasp force for the men was 1.6 times that of women, and this differencewas significant.

80

2 6 0

o-

*

Frmaleplanlsls

* --O-

llale Planlsls

--D-

l a l ec o m r o l s

Femaleconlrols

20

It Flnger Figure S-Mean groups.

pinch force of each finger for the female and male pianist and control

Aoki, Furuya,and Kinoshita

Discussion TappingAbility of ExpertPianistsVersusNonmusicians Single-Finger Tapping. Consistentwith our expectations,the trained pianists could tap their R and L fingers at much higher cadencethan the controls. This supportsthe notion that one major reasonfor the poor movem€ntability of these fingers in ordinary people is a lack of training resulting from limited use in a wide variety of manipulativedaily motor tasks(Kamakuraet al., 1980;Raj & Marquis, 1999). Conversely,the finding that the pianist-{ontrol differencein the I and M fingers is relatively small suggeststhat daily use of the hand itself can provide thesefingers with a fair amount of taining to maintain a near-maximumlevel of motor performance. The pianists had greater key-contactforce and impulse with the I and M fingers than the controls while therewas no group differencein theseforce values for the R and L fingers.Theseforce data,togetherwith the findings of the group difference in tapping cadence,suggestthat the level of force applied to the key did not seem to be related to the facilitation of tapping speed.Sternad,Dean, and Newell (2000) reportedthat the intraindividual variability of intertap interval was reducedwith an increasein impacting force during wdst tapping.The higher tapping force observedin our pianistsmight also have beenassociatedwith their intention to optimize thei timing accuracy.Indeed,ourpianistshad a smallervalue for intraindividual variability of intertapinterval for the I and M fingers. The tapping cadenceof the R and L fingers in thesetrained Pianistsdid not reach the level of the I or M finger The pianists were also perceivinggreaterdifficulty in tapping by the R and L fingers than the I and M fingers. The findings thereforesuggestthat even with extensivetraining of the individual fingers,some inherentfactors providing advantageouseffects of the I or M finger over the R or L finger cannot be overcomecompletely. What then haschangedafter extensivemovement training ofindividual fingers in pianists?First, one might haveto considertraining-inducedadaptivechangesin the functionalpropertiesof the musclesin the handandforearm.Our resultsshowed that the maximum isometric flexion force of any individual finger, as well as the maximum grasping force, was not pafticularly strongerin the pianistscompared with the controls. Therefore,muscle propertiesallowing strong force production do not seemto be a major sourcefor fastertapping ability of the pianists.There is evidencethat intrinsic handmusclesofpianists havea higher level of fatigueresistance againstcontinuoussubmaximalisometric force generation(Penn,Chuang, Chan, & Hsu, 1999). These researchershypothesizedthat piano training could induce improved capacity of aerobic function in the hand muscles via changes in metabolism,motor unit firing patterns,and muscle fiber composition.Because the presentrapid tapping taskswere performedonly for a short period of time and terminatedbefore muscularfatigue could be induced, the favorableeffect of this muscularproperty in pianistscould have had little effect on the presentoutcome. Second, the finding that the greatesttraining effect was found for the R finger indicatesthat neural and anatomicalmechanismsrestrictingisolatednnger

-

FinSer Tappingin Pianisrs

l5

movementcould havebeenchangedin the pianists(Hager-Ross& Schieber,2000; Zatsiorskyet al., 1998,2000). In the studiesof force interactionamons the finsers in isometricforce productionrasks,so-calledenslavingeffects,ZatJiorskyei al. (1998,2000) discussedthe underlyingmechanismsthat could makeisolatedfiager movement difficult. These included peripheral tendinous interconnections,ahe motor units that are locat€dwithin a neuromuscularregion servingone finger also insert into another finger, and divergenceand convergencein central commands. Presumably,some or all of thesemechanismshave changedin the €xpertpianists to conquerthe difficulty ofmoving individual fingers separately. Tendonsconnectedto multiple fingers are interconnectedby tendinousor fascialike structuresin the extensordigitorum communis (EDC) and flexor digitorum profundus (FDP) muscles (Fabrcr, l98l; Von Schroeder& Botte, 1993), which producecoupling movementsin the fingers adjacentto an intendedfinger. Use- or training-dependentchangesin this anatomical structure are, nowever, less likely to occur. This could partially exptain the finding that gr€atertask difficulty with the R finger movementhas continuouslybeenexperiencedeven after years of training in the pianists.The secondmechanismconcemsthe Dresenceof functionally independentneuromuscularregions(companments)within the same muscle that have connectionswith multiple fingers. Schieber,Gardinier,and Liu (2001) demonstratedthat electrical stimulation applied to each of theseregions in monkey FDP muscle produceduniquely distributed tension on all five digits. More recently, Keen and Fuglevand (2004) reported the presenceof synchronized activity in motor-unitpairs in neighboringcompartmertsof the humanEDC muscle during low-force isometric extensionof all four fingers. They also found that the largest value of intercompartmentsynchrony was for the R-finger motor units and the least synchony was for the I-finger units. The third mechanismis neural interaction at the central level. It is now well known that there are convergence and divergenceof output from the primary motor cortex for the activationof hand muscles(McKiernan, Marcario, Karrer, & Cheney, 1998; Uematsu,et al., 1992). The extent of plasticity with movementtraining in theseneural interactionsboth at the pedpheral and central levels,however,remainsunknown. It is then possible to assumethat training-related changesin spinal and supraspinalneural mechanismscould have principally contributed to the higher level of motor performancein the pianists.Many studieshayeshownfunctionaland structuralchangesin the motoneuronfiring threshold,axonalconductionvelocity, and synapticterminals on motoneuronsfollowing repetitivelong-term and intensive practice for acquisition of motor skills (seereview by Wolpaw & Tennissen, 2001). The supraspinalneural mechanismsare mainly motor-aisociatedcortical and subcortical shuctures,which will be discussedin relation to the findines of double-finger tapping. Double-Finger Tapping. The presentdouble-fingertapping experimentdemonstratedsignificant pianist-{ontrol differencesin the mean values of all temporal parametersfor eachfinger-combination.The differencewas relatively small when performing the easiesttask of the IM combination, but ir was fairlv larse for the most difficult task of the RL combination. This occurred becauserhJ pranrsts

-

Aoki,Furuya, and Kinoshita performed all finger-combinationtasksrelatively quickly with a small decline in cadencefrom the IM to RL combinationswhile the controls were extremelyslow when performing the RL combination. Our recent PET study using young nonmusician participants has shown that the RL combination task demandsinvolvementof much sreatercortical and subcorticalareaswith a higher level of activity in each areaiompared with the IM combination or single-fingertapping by the I or R finger (Aoki et al., 2004). The finding that the pianists could tap the RL combination at relatively similar tapping cadenceto that for the IM combination might imply decreaseddemand on cental neural function when the pianistsperform the RL combination.Indeed, use-dependentcortical and subcortical reorganizationfor efficient use of brain function is now a well-known phenomenon(Doyon, Penhune,& Ungerleider, 2003). Neuroimagingstudiesdemonstratedthat activebrain areaswere reducedin highly practicedand automatedmoyementcontrol, requiring little overt attention. For example,using a functional magneticresonanceimaging (fMRI) technique, Jancke,Shah,and Peters(2000) demonstratedthat during the performanc€of selfpaced bimanual and unimanual tapping tasks,the primary and secondarymotor (supplementaryand cingulatedmotor) areaswere activatedto a much lesserdegree in professionalpianiststhan in nonmusicians.Basedon thesefindings,the authors stated that long-term hand skill training leads to greater efficiency, which was reflectedin a smaller number of active neuronsneededto perform a given finger movement.Hund-Georgiadisand Von Cramon (1999), who examineda complex finger-thumb opposition task, reporteda similar finding. The ability to move the individual finger(s)at a fasterspeedby the pianists,therefore,most likely reflected an increasedcapacity of their central neural function. In addition, the central program to actlvate appropriat€ setsofmuscles with apFopriate intensity and optimized coordination for the flexion-extensionmovementof the R or L fineer could have beenestablished with yearsof pianoraining. Intraindividual Variability. Temporal and force accuraciesare essentialcomponents in piano performance,and thereforethe trained pianists are expectedto demonstratea lower degree of intraindividual vadability in intertap interval, as well as in key-striking force, than nonmusiciancontrols.The resultsofthe present study partially supportedthis hypothesis.The pianists indeed had smaller mean CV valuesofintertap interval for the I and M fingers in single-fingertapping than the controls,confirming higher temporalcontrol ability in the pianists.In contrast, the correspondingCV valuesfor the R and L fingers and all three pairs in doublefinger tapping were not different betweenthe two groups. The previous studies demonstratedthat the CV value of the intertaDinterval was relativelv small and constant(Collyer,BoatrighrHorowitz.& Hooper,1997:Peters,1989)or slightly decreased(Sternadet al., 2000) from the changesin cadenceat l-3 Hz; nowev€r, it increaseddisproportionatelyoutsidethis range(Peters,1989).The meanvalues of tapping cadencein all tapping conditions in the presentstudy exceeded3 Hz, and therefore, the effect of tapping cadenceon amplifying the CV value might have been greater for the pianists who had a higher cadencecompared with the controls. The inability to lind a significant group difference in the CV values of

-

FingerTapping in Pianists

37

the R and L fingers or all finger-combinationsmight have been attributable to this eff€ct. There was no pianist-+ontrol differencein the CV valuesof averageforce in any ofthe fingersor finger-combinations.A discrepancyin the outcomeof temporal and force variability appears to support the notion that the control of force and the control of timing are independent(Keele,Ivry, & Pokomy, 198?)and that they demanddifferent levels ofactivation in the cortical and subcorticalsrrucrures (Wexler et al., 1997).The CV valuesofaverageforce were found to be much larger than thoseofthe intenap interyal in all experimentalconditions.This is consistent with the findings of the previous study in which the CV value of peak tapping force variability during wrist tapping was approximately207o,whereasthat of the intertapinteryal was approximately57o(Stemad€t al., 2000).The finding that the pianistswere not particularly superiorin producinga constanttapping force might be attributedto the lack of feedbackinformation, b€causethey commonly practice with soundfeedback.In this regard,the pianistsand conhols were equally naiveto the presenttask.A more detailed study is neededto inyestigatethe effects of skill level on force and timing variability during tapping movements.

Cender Difference A genderdifferencewas found for the isometric maximal-force-productiontasks, but not in any of the presenttapping tasks.Somestudiesshoweda male advantage in tapping cadence(Nicholson & Kimura, 1996; Peters& Campagnaro,1996; Ruff & Parker, 1993) while some reported no genderdifference (Hermsdorferet al., 1999;Mckeever& Abramson, 1991).Our findings are, therefore,in agreement with those of the latter studies.Common explanationsfor the hypothesizedmale advantageare that men have greater muscular strength,a higher level of fatigue tolerance,and exhibit more overt competitivenesswhen pedorming motor tasks. Our force measurementdata support the idea of the male advantagein muscular strength.In addition, our subjectiveimpressionthoughout the testing also supported the idea of greatermotivation in male participants.Nevenheless,our data indicated that female panicipants w€re as fast as male participantsin ali rapprng tasks. Severalreasonsfor the lack of gender difference are conceivable.One is that the fingers are used constantlyin our everydayactivities for both gendersin pianists and conhols, and thus they are nearly equally trained and optimized to a level adequatefor daily use.Second,althoughmuscle strengthis greaterin mal€s than in females,this adyantagecould be offset by a larger finger mass in males and also by their mor€ pronouncedenslavingeffects (Shinoharaet a1.,2003).In addition, tapping movementby the individual fingerscan be a sensorymotor task with a higher skill elem€nt, requiring the involvement of greater central neural function than the muscle-powerelement.

Acknowledgments This study was supportedby a SasakawaScientific ResearchGrant of the Japan ScienceSociety.

-

Aoki,Furuya,and Kinoshita

References Aoki, T., Francis, PR., & Kinoshita, H. (2003) Differences in the abilities of individual fingers during the performance of fasi, repetitive tapping movemefis. Expcrimental Brain Research,152. 2'70-280. Aoki, T., & Kinoshita, H. (2001).Temporal and force characteristicsof fast double-finger, single-finger and hand tapping.Ergonomics,44, I 368-1383. Aoki, T., Tsuda, H., Takasawa,M., Osaki, Y, Oku, N., Hatazawa,J., & Kinoshita, H. (in press).The effect of tapping finger and mode differenceson cortical and subcortical activities:A PET study.Expeimental Brain Research. Collyer, C.E., Boatright-Horowitz, S.S., & Hooper, S (1997). A motor timing exPeriment implementedusing a musicalinstrumentdigital interface(MlDl) apptuach.Behavior Research Methoh, Insttuments, and Computers, 29, 346-352. Doyon, J., Penhune,V, & Ungerleider L.G. (2003). Distinct contributiod of the corticostriatal and cortico-cerebellar systems to motor skill lea'x,ing. Neuropsychologia' 41.252-262. Fahrer, M. (1981). Interdependentand independentactions of the fingers ln R. Tubiana (Ed.), The hand (pp.399-403). Philadelphia:Saunders Hager-Ross,C., & Schieber,M.H. (2000). Quantifying the independenceof human finger movements: Comparisonsof digits, hands, and movementfrequencies.Joumal of N euroscience. 20. 8542-8550. Hermsdorfer, J., Marquardt, C., Wack, S., & Mai, N. (1999) ComParativeanalysis of diadochokinetic movements. Journat of Electromyography and Kinesiology' 9'

283-29s. Heyward,VH., Johannes-Ellis,S.M., & Romer,J.F.(1986).Genderdifferencesin strength. ResearchQuarterlyfor Exerciseand Sport,57, 154-159. Hund-Georgiadis,M. , & Von Cramon,D.Y ( 1999).Motorleaming-related changesin piano players and non-musicians revealedby functional magnetic-rcsonancesignals. Erperimental Brain Research,125. 417-425. Jancke,L., Shah, N.J., & Peters,M. (2000). Cortical activationsin primary and secondary motor areas for complex bimanual movementsin professionalPianists.CoStitive B rain Research.10. 177-183. Kamakura, N., Matsuo, M., lshii, H., Mitsuboshi, F., & Miura, Y (1980). Pattemsof static prehensionin normal hands.American Joumal of Occupational Therapy,34, 431M5. Keele, S., lvry, R., & Pokomy, R. (1987). Force control and its relation to timing. ./orrna, oJMotor Behavior 19,96-114. Keen, D.A., & Fuglevand,A.J. (2004). Common input to motor neurons innervating the sameand different compartmentsof ihe human extensordigitorum muscle Joumal of Neurophysiology,91, 57-62. Kinoshita, H., Murase, T., & Bandou, T. (1996). Grip posture and forces during holding cyfindrical objects with circular grips. Ergonomics,39, 1163-l176. Mckeever.WF. & Abramson,M. (1991).Halsteadand Halstead-Reitannorms for the linger tapping test are severelybiasedagainstfemalesand left-handers Joumal ofClinical and Etperimental Neumpsychology,13, 91. McKieman, B.J., Marcario, J.K., Karrer,J.H., & Cheney,PD (1998) Corticomotoneuronal

-

FingerTapping in Pianists postspikeeffectsin shoulder,elbow, wrist, digit, and intrinsic hand musclesduring a reachand prehensiontask.Jo mal of Neurophysiotogy,80,I961- 1980. Nicholson, K.G., & Kimura, D. (1996). Sex differencesfor speechand manual skill. perceptual and Motor Skills,82,3-13. Oldfield, R.C. (1971).The assessment and analysisof handedness: The Edinburghinventory Neuropsychologia,9 , 97-l 13. Penn, I.W, Chuang,T.Y, Chan, R.C., & Hsu, T.C. (1999). EMG power spectrumanalysis of first dorsal interosseousmuscle in pianists.Medicine and Sciencein Sportsand Exercise,31. 1834-1838. Peters,M. (1989). The relationshipbetween variability of intertap iniervals and interval dtratiotr. Psychological Research, 51, 3842. Peters,M., & Campagnaro,P (1996). Do women really excel over men in manual dexterrly'l loumal of Erpeimental Psychologf: Human perception and perJormance, 22,

tto7-1112. Radwin,R.G.,Oh, S.,Jensen, T.R.,& Webster, J.G.(1992).Extemalfingerforcesin submaximalfive-fingerstaticpinchprehension. Ergonomics, 35,275-289. Raj,R., & Marquis,C. (1999).Fingerdominance. Joumalof HandSurgery,A,429-430. Ruff, R.M.,& ParkeaS.B. (1993).Gender-andage-specific changes in motorspeedand eye-handcoordinationin adults:Normativevaluesfo. thefrngertappingandgrooved pegboardtests.Perceptualand Motot Skilk, 76, 1219-1230. SchieberM.H., Gardinier, J.,& Liu, J. (2001).Tensiondistributionto rhefivedigitsof the handby neuromusculalcompartmentsin the macaqueflexor digitorum plofundus. Joumalof Neuroscience, 21,2150-2158. Shinohara, M., Li, S., Kang,N., Zarsiorsky, VM., & Latash,M.L. (2003).Effectsof age andgenderon nngercoordination in MVC andsubmaximal force-matching tasks. Joumal oJApplied Physiology,94,259-27O. Staron,R.S.,Hagerman, F.C.,Hikida,R.S.,Munay,T.F.,Hostler,D.p, Crill, M.T.,et al. (2000).Fiber type compositionof the vastuslateralismuscleof youngmen and rr/omen.Joumal of Histochemistryahd Cytochemistry,48,623-629. Stemad,D., Dean,WJ., & Newell,K.M. (2000).Forceandtimingvariabilityin rhythmic unimanualtapping.Journalof MotorBehavioa32,249-267. Uematsu, S.,l,esser, R.,Fisher,R.S.,Gordon,8.,Hara,K., Krauss,C.L.,et al.(1992).Motor andsensory cortexin humans: Topography studiedwithchronicsubdural stimulation. Neurosurgery, 31, 59-1l. VonSchroeder, H.P,& Botte,M.J.(1993).Thefunctionalsignificance of thelongextensors andjuncturae tendinumin fingerextension. Joumalof HandSurgery,18, 641_647 . Wexler,B.E.,Fulbright,R.K.,Lacadie,C.M.,Skudlarski, p, Kelz,M.8., Constable, R.T.. & Gore,J.C.(1997).An IMRI studyof thehumancorticalmotorsystemresponse to increasing functionaldemands. MagneticResonance Imaging,lS, 385-396. Wolpaq J.R.,& Tennissen, A.M. (2001).Activiry-dependenr spinalcordplasricityin health anddisease.Annual Reviewof Neurcscience,?A,80'l-843. Zatsiorsky, VM., Li,2.M., & Latash,M.L. (1998).Coordinated forceproductionin multifingertasks: Fingerinteraction andneuralnetworkmodeling. Biological Cybemetics, 79,139-150. Zatsiorsky, VM., Li, 2.M., & Latash,M.L. (2000).Enslavingeffectsin multi-fingerforce prodttction. Erperimental Bruin Research,l3l,187-195.