- Email: [email protected]
Bereitschaftspotential during the acquisition of a skilled motor task
568
Electroencephalography and Clinical Neurophysiology , 1978, 45:568--576
© Elsevier/North-Holland Scientific Publishers, Ltd.
B E R E I T S C H A F T S P O T E N T I A L D U R I N G T H E ACQUISITION OF A S K I L L E D MOTOR TASK MARGO J. TAYLOR 1 Simon Fraser University, Burnaby, B.B. (Canada)
(Accepted for publication: March 17, 1978) Although movement-associated cortical potentials in man were first r e p o r t e d by Bates in 1951, it was n o t until 14 years later t h a t the slow negative wave preceding voluntary m o v e m e n t , the Bereitschaftspotential (BP), was first discovered and named ( K o r n h u b e r et al. 1965). Since then much research has focused on the role of the responding musculature and o f m o v e m e n t parameters on the size and cortical distribution of the BP. Research t h a t is c onc e r ned with more conative aspects of m o v e m e n t , or t hat does n o t view the BP as simply an electrophysiological correlate of the m o v e m e n t itself, is less plentiful. Becker et al. (1976) r e p o r t e d slow movements to be preceded by BPs of greater amplitude and duration than were ballistic movements. T h e y speculated t h a t m or e preparation was required for the slow, controlled m o v e m e n t s and this interpretation has been supported by o th e r researchers. Loveless and Sanford (1974) r e p o r t e d the amplitude of the wave preceding a response to be p r o p o r t i o n a l to the level of p r e p a r a t o r y set, as inferred f r o m the speed o f the reaction time. McAdam and Rubin (1971) used response accuracy as a measure o f preparation to respond and f o u n d larger BPs with accurate responses. Their subjects' post-response estimates of accuracy were positively related to the actual accuracy obtained. Ford et al. (1973) investigated BPs with qualitatively di f f er e nt b u t t o n presses. Pressing a skin-contact b u t t o n , which required more preparation than a standard push-
1 Present address: Psychology Department, McGill University, Montreal, Que., Canada.
b u t t o n , was preceded by BPs of increased amplitude and duration. McCallum ( 1 9 7 6 ) 2 report ed that as the level of task involvement increases, corresponding increases were f o u n d in the BP, with frontal areas making relatively greater contribution. A c o n f o u n d i n g variable in the study of Ford et al., however, is post-response feedback or stimulation. Their subjects received more sensory feedback concerfling the accuracy o f their response from the skinc o n t a c t b u t t o n than from the standard but t on. It has been well d o c u m e n t e d that feedback from or contingencies placed upon the m o v e m e n t affect the BP amplitude and duration (McAdam and Seales 1969; Dincheva and Hardin 1975; McCallum 1976 2; T ayl or 1976 3). Research to date with BPs has used simple repetitive movements, with little a t t e m p t to investigate the d e v e l o p m e n t of a specific m o v e m e n t or skill. Only one study has been report ed in which the voluntary response could be considered a skilled m o v e m e n t (Papakostopoulos, 1976 4). The skilled task was a single b u t t o n press, the timing of which was to be learned. Papakostopoulos f o u n d the BPs preceding correct responses were larger 2 McCallum, W.C. Relationships between the Bereitschaftspotential and the CNV. Paper presented at the 4th International Congress on ERSP, Chapel Hill, N.C., 1976. 3 Taylor, M.J. BP amplitude as a function of response contingencies. Paper presented at Society for Psychophysiological Research, San Diego, Calif., 1976. 4 Papakostopoulos, D. Electrical activity of the brain associated with skilled performance. Paper presented at the 4th International Congress on ERSP, Chapel Hill, N.C., 1976.
BP AND SKILLED TASK ACQUISITION than those preceding incorrect responses and BPs preceding the skilled task were larger than those preceding the unskilled task. Although this study offers further support for the thesis t h a t BPs reflect the subject's preparation for and accuracy in responding, problems arise with the definition of a skilled task. There was no increase in the number of correct responses over the session, y e t Papakostopoulos discussed the results in terms of developing and improving skillful performance. His subjects were n o t learning a response as much as selecting the response on cue. Research in the field of m o t o r skills, however, has shifted from an emphasis on product, or the selection of a response, to an emphasis on processes occurring during the learning of a skilled response (Trumbo and Noble 1973). For better integration of BP research with the field of m o t o r tasks, a response which can show clear improvement in performance with practice should be chosen as the voluntary movement. The m o t o r skill literature also offers much information on other aspects of a skilled task that should be considered in this type of research. The performance of a task gives rise to several sources of feedback and recently much interest has centered on the role of feedback in skilled movements. When subjects receive feedback from proprioception and vision, they are able to accurately estimate the precision of their response after the m o v e m e n t and correct errors on subsequent trials (Adams 1971, 1976; Schmidt 1976). Under conditions of ample feedback a skilled movement can rapidly approximate the response described in an instructional set; the most effective instruction to achieve this end would be one emphasizing speed (Bouisset and Lestienne 1974). As elements are added to a serial task the period required to master it lengthens and attentional requirements for the response increase. Klein and Posner (1974) report that simple discrete movements demand no attention except at initiation, but that attentional demands t h r o u g h o u t a serial m o v e m e n t
569 increase with the level of accuracy required. Improved performance is accomplished only with the allocation of additional attention (Klein 1976). The attentional demands of a repetitive serial task diminish only after the sequence is learned and response uncertainty decreases. Also, a movement t h a t a subject is prepared to make involves more attention and is performed better than an unanticipated movement. All this suggests that attention is related to preparation to respond and accuracy of the response, with obvious implications for BP research. Research in the area of m o t o r skills often parallels the research with BPs, y e t few studies have made these parallels explicit. Non-skilled self-initiated movements have been well studied by BP researchers, as have m o t o r skills by m o t o r performance, control and learning theorists. Research that can interrelate knowledge from both fields is necessary if a fuller understanding of the mechanisms involved in m o t o r skills is to evolve. This study was designed to investigate the BP and its cortical distribution during the acquisition of a skilled serial m o t o r response. Improvement in the speed of the response over trials was established as the measure of skill acquisition. A standard condition was conducted both before and after the skilled task condition to determine if any systematic changes occurred in the BP as a function of the duration of the experimental session. It was expected t h a t the BP would increase in size during learning, particularly over the frontal areas, and then diminish after acquisition. The initial rise in the BP was expected as response certainty increased; the frontal area was expected to change more rapidly to reflect preselection or the anticipatory planning of the m o v e m e n t (Teuber 1964; Kelso and Stelmach, 1976).
Method
Sub]ects Six male and 6 female paid subjects were
570
recruited from the psychology department at Simon Fraser University, ranging in age from 19 to 31 years. All were right-handed and all were naive as to the purpose of the study. Also none had had previous experience in slow potential research.
Apparatus Non-polarizable Beckman silver-silver chloride electrodes were affixed to the subject's scalp with collodion soaked gauze at Cz, Fz and two lateral placements, 5 cm lateral and 2 cm posterior to Cz (C3" and C4") (Papakostopoulos 1976). An infraorbital EOG and the EEG electrodes were referred to linked mastoids. The impedance between any two electrodes was less than 5 k~2 for all subjects. The EEG signals were amplified by Grass 7 P I A DC amplifiers, with a roll-off of 3 dB at 50 c/sec. The single trials were collected and digitized (1024 points per sweep) by a Hewlett-Packard 2116B computer. Data were collected from the ongoing record during a period 4 sec prior to and 2 sec following the initiation of the subject's response. In case of contamination by an artifact, such as eye movement, the trial was rejected and the subsequent artifact-free trial accepted in its stead.
Experimental procedure The subjects were seated in a large comfortable chair in an electrically shielded room and given a small metal box (12.5 cm × 10 cm × 8 cm) which they held on their laps. The box contained 2 rows of 3 buttons each, with 2 cm between adjacent buttons. A force of 1050 g was required to depress the buttons the necessary 5 m m for switch closure. Subject's right arm rested on the arm of the chair and was supported with a pillow to minimize involvement of the forearm when responding. All subjects used only the index finger of their right hand throughout. They were instructed to refrain from blinking and moving their eyes, particularly during the few seconds before and after a response. Subjects were asked to respond approximately every
M.J. T A Y L O R
20 sec, but were requested n o t to count or use a watch as some variability in timing was desirable. Between trials, the subject's finger rested on the first b u t t o n ensuring that the first m o v e m e n t initiating the trial was identical across trials and conditions. The first or standard condition required the subject to make a single b u t t o n press. The subjects were instructed to respond sharply as they would in a reaction time task. During this condition the subjects were given feedback, via an intercom, regarding spacing their response (i.e., whether there was either too little or too much time between trials) and eye movements (whether they were blinking or moving their eyes during trials). SubjecSs generally required very little feedback during the session. After collecting 15 trials of the standard condition instructions for the experimental condition were given. A response for this condition involved pressing a series of buttons in rapid succession. The subject was shown the pattern of b u t t o n presses, which included all 6 buttons with no repetitions (Fig. 1). The pattern was demonstrated 3 times by the experimenter but subjects were allowed no practice. Subjects were instructed to press the series of buttons as quickly as possible, always returning to the first b u t t o n upon completion. The requirement of maximizing speed w i t h o u t sacrificing accuracy was stressed. This
5
3
4
6
1
Fig. 1. The pattern of button presses constituting the serial task for the skilled task condition.
BP A N D S K I L L E D T A S K A C Q U I S I T I O N
571
emphasis led to the desired result of subjects making no errors at all in the pattern of b u t t o n presses but marked improvement over trials in the speed with which the series was pressed. Subjects were told t h a t the spacing of the responses and control of eye m o v e m e n t was to be the same as in the preceding condition, but that the duration of the experimental condition would be much longer. Forty-five trials were collected. The subjects were then informed that the standard condition was to be repeated. The earlier instructions reiterated and a final 15 trials collected. The entire experiment, including electrode application, t o o k no more than 1.5 h. Although single trial analysis would have been ideal in this study, some averaging before analyses were conducted was necessary due to the noise in single trial EEG data. In the experimental or skilled task condition, data from sets of 5 consecutive trials were averaged, yielding 9 averages for each channel for each subject. The first second of data from each trial was taken as baseline before averaging. Algebraic area measures were calculated over
3 sections of each average. The sections were defined as follows: (1) the 2 sec period prior to the response; (2) minus to plus 50 msec from the response, a peak measure; (3) the 2 sec period following response initiation. The duration of the response from the first through sixth series of b u t t o n presses and the time between the first and second b u t t o n presses were measured for all 45trials. Response time over the series of 6 presses was averaged in sets of 5 to correspond with the EEG averages for the analyses involving both measures.
Results An analysis of variance was conducted for each of the BP measures to test for difference among the two standard and the skilled task conditions. No significant condition main effect was found for the BP measure 2 sec prior to the response (F(2,20)= 1.39, NS), whereas the peak measure was significantly larger during the skilled task than during the sec
Cz
c3"
c~.
EOG
?
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
Fig. 2. BPs averaged across subjects for each e l e c t r o d e for the 9 trial blocks of the skilled task (each wave f o r m is a c o m p o s i t e of 60 single trials; o n e trial block of 5 single trials from each of 12 subjects).
572
two standard conditions ( F ( 2 , 2 0 ) = 4 2 . 7 8 , P < 0.001). The averages across the subjects for the skilled task, for the 9 trial blocks and 4 electrodes, illustrate the general trends in the data (see Fig. 2). The influence of skill acquisition on the BP was investigated using an analysis of variance for each of the dependent area measures. Significant trial (F(8,80) = 2.34, P < 0.05) and electrode (F(3,30) = 12.64, P < 0.001) main effects and a trial by electrode interaction (F(24,240) = 1.91, P < 0.005) were found for the area measure 2 sec prior to the response and the peak area measure (respectively, F(8,80) = 2.75, P < 0.025; F(3,30) = 17.02, P <: 0.001; F(24,240) = 2.76, P<0.001). A significant electrode main effect was found for the area measure 2 sec following response initiation (F(3,30) = 8.94, P < 0.001). No significant change over the trial blocks was found for this measure ( F ( 8 , 8 0 ) = 1 . 2 2 , NS). The electrode main effects were attributable to the constant amplitude differences among the electrodes. Across subjects and trials the size of the BP increased from Fz, to C4", to C3", to C~. The greatest difference in the BP measures was between Fz and the other 3 electrodes (F(1,30) = 20.36, P < 0.001). The BP at C3" was n o t significantly larger than at C4" (F(1,30) = 3.85, P < 0.1) although the trend was in the expected direction. For the last area measure, 2 sec following response initiation, both Fz and C4" were of much smaller area than were Cz and C3" (F(1,30) = 25.86, P < 0.001). The trial main effects were due to a steady increase in the area, at all electrodes, for the first 4 averages (or 20 trials) and then some decrease, on average, over the next 4averages (or 20trials). The interaction effects were due to differences between the electrodes during the last 5 averages. At C4" and F~ the BP decreased steadily from 3 to 4 of these last 5 averages while at C~ and C3" the size dropped initially and then rebounded for 1--2 averages. At all electrodes there was an increase in the size of the BP during the last average (see Fig. 3 and Table I).
M.J. T A Y L O R -22
0 -Cz - FZ
.,,Q
•.-~..
........
O -C4""
~"
\ i
\ ....:~ \
/ ~ .....
U
...:"
,,,l
%; , -'.
:~ ......
[ ~
/ -/. :
\ ' "...
.~..
,~"
~" -
.:
? ......... ~
!,.
"a.
/""
/ - ,.
,~.. . .~
Y ." ......-"
.....
"".,~ .....
./
.,~. ......... ,,~:
..."
TRIAL
BLOCKS
Fig. 3. BP area m e a s u r e 2 sec p r i o r to t h e r e s p o n s e averaged across s u b j e c t s for t h e 9 trial b l o c k s o f t h e skilled task c o n d i t i o n .
The response time over the series of 6 b u t t o n presses decreased for all subjects over trials (F(8,80) = 28.42, P < 0.001) reaching asymptote at about the 20th trial or 4th average (Fig. 4). The response times between the first and second b u t t o n presses over the 45 trials follow the same pattern as the total response times (Fig. 5). The EEG data were also analyzed using a multivariate analysis of covariance with the
TABLE I BP area m e a s u r e s 2 sec p r i o r t o t h e r e s p o n s e for s t a n d a r d c o n d i t i o n s a n d skilled task trial blocks.
Standard Skill task Skill task Skill task Skill task Skill task Skill task Skill task Skill task Skill task Standard
I TB1 TB2 TB3 TB4 TB5 TB6 TB7 TB8 TB9 II
Fz
Cz
C3"
C4"
--6.43 --1.83 --5.08 --8.16 --10.86 --8.42 --6.50 --6.00 --5.68 --9.57 ---6.46
--12.10 --8.17 --11.26 --18.98 --20.85 --12.53 --17.51 --17.29 --11.29 --19.82 --14.32
--11.62 --8.32 --12.17 --17.40 --16.28 --12.09 --16.18 --11.76 --12.38 --14.89 --11.55
--8.05 --4.53 --11.64 --12.17 --16.47 --11.51 --9.52 --8.33 --9.62 --11.98 --9.49
BP A N D S K I L L E D T A S K A C Q U I S I T I O N
573
35"
sistent relationship between the changes in the BP and the improvement in the response times over trials {Sprott 1970).
Discussion
~,t,,,B,,,l,l,,i,,,I,l,,ss,l,,
,,*l~l,t,,,l**4,
TRIALS
5
Fig. 4. Average r e s p o n s e t i m e f r o m t h e first t h r o u g h s i x t h series o f b u t t o n presses for t h e 45 trials of t h e skilled task.
response time as the covariate. No significant trial main effects were found: F~ (F(32,281) = 1.06, NS), Cz ( F ( 3 2 , 2 8 1 ) = 1.26, NS), C3" (F(32,281) = 1.60, NS) and C4" (F(32,281) = 1.25, NS). Thus, when the EEG data across trials were adjusted for the variance contributed b y the response times the trial main effect d r o p p e d out. This demonstrates a con-
°81 ~,( •
•
~'~"
"~.~fl
"~.~'A
o•
w
TRIALS
Fig. 5. Average r e s p o n s e t i m e b e t w e e n t h e first a n d s e c o n d b u t t o n presses for t h e 45 trials o f t h e skilled task.
The results offer strong support for the hypothesis that the size and cortical distribution of the BP change with the acquisition of a skilled m o t o r task. The magnitude of the BP at all electrode locations increased steadily over the trials in which the response time was decreasing, i.e., during the acquisition of the response. The two essential requirements of the study were found for all subjects. First, a clear learning effect of the m o t o r skill and second, no significant differences between the two standard conditions, indicating that systematic changes in the BP were n o t a function of the duration of the experimental session. .As subjects acquire a skilled response, their response accuracy should improve. An increase in accuracy and preparation to respond has already been related to increased size of the BP (McAdam and Rubin 1971; Ford et al. 1973; Loveless and Sanford 1974; Papakostopoulos 1976). In all previous research, however, the measures of certainty, accuracy and preparation have been dichotomized. The subject's responses were n o t learned b u t selected and then performed either correctly or incorrectly. With the learning of a m o t o r skill gradual changes in certainty and preparation would be expected to parallel gradual improvement in performance. Subjects in the present study were fully aware of the increase in the precision of their performance over the experimental session. Thus, the systematic increases in the BP area, concurrent with the steady improvem e n t in response performance, add further credibility to the theory that the BP reflects preparation and certainty of the subject to respond. It was expected that the area of the BP 2 sec prior to the response would be larger
574 preceding the skilled response than before the single b u t t o n press. The demands placed upon the subject in terms of preparation, preselection and attention during a skilled task far outweigh those necessary for a simple movement. Also, greater amounts of information must be monitored for the performance of a skilled task compared to an unskilled task; an increase in post-response feedback has been shown to increase BP amplitude and duration (Dincheva and Harding 1975). A difference between the skilled and unskilled tasks, however, was found only for the peak measure, which does correspond to the literature which has employed only a peak amplitude measure. The explanation, though, for this lack of significance with the 2 sec BP measure may rest in the large variations in the size of the BP within the skilled task condition. For the first one or t w o initial trial blocks of the skilled task the BP was smaller at all electrode placements than it was during the standard conditions. On subsequent trials it was larger than the standard conditions except for one or t w o trial blocks after learning (see Table I). According to the hypothesis presented in this study these results would imply less accuracy and preparation for the performance of the response during the initial 5 or 10 trials of the skilled task than for the standard response. When the skilled response is acquired and uncertainty decreased, the corresponding BP is larger than in the standard conditions. The decline on some trials after learning could be attributable to decreases or lapses in the subject's attention. Klein (1976) reports attentional demands of a serial task to diminish after the task sequence has been learned. The significantly larger peak measures in the skilled task condition appeared, from visual inspection, to be concomitants of increased negativity of slope of the BP preceding skilled movement. In the standard conditions the slope of the BP was usually linear, in accordance with previous studies which employed simple movements. In the skilled task condition the slope was quadratic
M.J. TAYLOR or cubic. From this study it is impossible to determine whether the apparent difference in slope between the conditions is a function of the learning required for the skilled task or the duration of the skilled task. It was expected that the role of the frontal areas in organizing m o t o r outp.ut and in monitoring o u t p u t and feedback would be reflected in qualitatively different changes in the frontal BPs than those over the m o t o r cortex. The electrode by trial interaction, however, is due to differences only after the response had reached asymptotic levels. The frontal BPs increased at the same rate as the vertex and lateral BPs during acquisition. After response acquisition, the frontal BP, in contrast to those from Cz and C3", decreased steadily. With practice the performance of a response tends to become automatized lessening both attentional demands and the a m o u n t of feedback requisite for maintenance of the response (Keele and Summer 1976; Klein 1976). Probably the subject's involvement in the task also decreased after the goal of fast, accurate responses had been achieved, which would predict a reduction of frontal BPs (McCallum 1976). The relative size of the BPs in relation to their cortical distribution is congruent with most other research in the field. The BP was largest at Cz, larger over the m o t o r cortex than frontal areas and tended to be larger contralateral to the responding musculature. Although the peak measure corresponds to the m o t o r potential (MP) of other researchers (Vaughan et al. 1968; Deecke et al. 1973), it did n o t follow the same pattern of lateralization. As earlier research averaged a very large number of trials before the MP was evident, this difference in methodology alone could account for the lack of a lateralized peak measure in this study. The divergence between C3" and C4" after learning (see Fig. 3) suggests that the less attention a movement requires the more specific to the m o t o r projection area the BP becomes. In concurrence with this, McAdam and Rubin (1971) suggested that their finding of non-
BP AND SKILLED TASK ACQUISITION lateralized BPs was a function of the attentional demands placed u p o n their subjects. This offers an explanation of the maintenance of large BPs close to the m o t o r projection area after acquisition when BPs recorded from more remote locations decreased. Continued negativity after the initiation of the response is n o t generally reported in the literature, as a large positive wave usually follows the response within 200--300 msec. This discrepancy between the present study and the existing literature is likely due to the duration of the serial response. During the 2 sec following response initiation the response is still being performed; as can be seen in Fig. 4 the average response time, even after learning, was over 1.5 sec. In other BP research the positive wave follows the initiation of the response, b u t the responses have been simple short movements, such that it also follows the termination of the response. As the subjects acquired the skilled task and the response duration shortened to less than 2 s e c this post-BP negativity tended to decrease, although the effect was n o t significant. In this study, however, the BP reflects the level of m o t o r skill learning and, after learning, possible changes in the attention devoted to the task (Klein 1976). If one is interested in the mechanisms involved both during and after the acquisition of m o t o r skills, the BP could add a valuable dimension to the research. Potentially, it could be used to distinguish between movements that require central attentional mechanisms for initiation and performance and those that do not. The data suggest a causal relationship although, from this study, one cannot claim that the BP is an essential c o m p o n e n t of m o t o r skill acquisition, b u t certainly the question is posed for future research.
Summary Bereitschaftspotential (BP) research typically uses very simple, abrupt movements
575 that require no learning and which change little over the duration of the experimental session. The present study investigated changes in the size and cortical distribution of the BP during the acquisition of a skilled m o t o r task. Twelve subjects were employed. Electrodes at Fz, Cz, C3" and C4" were used to record the EEG with DC amplifiers. A series of 6 b u t t o n presses in a specified pattern constituted the m o t o r task. Subjects were instructed to press the series every 20 sec as quickly as possible, b u t with no errors. Significant response time, electrode and trial main effects, and electrode by trial interactions were found. The BP increased steadily at all electrodes as performance improved, i.e., as response time decreased. After the response reached asymptote the BP recorded at Fz and C4" decreased, while the BP at Cz and C3" remained relatively constant. Multivariate analyses of covariance showed a consistent relationship between the improvement in response times and the changes in the BP. This study demonstrates that the size and cortical distribution of the BP are systematically related to improved proficiency of a m o t o r response with learning.
Rdsumd Potentiels de prdparation motrice au cours de l'acquisition d'une tdche motrice complexe
Les recherches sur le potentiel de preparation motrice (BP) utilisent d'ordinaire des mouvements tr~s simples et brusques qui ne n6cessitent aucun apprentissage et qui se modifient peu sur toute la dur~e de la session exp~rimentale. Cette 6tude recherche les modifications de dimension et de distribution corticale du BP au cours de l'acquisition d'une t~che motrice complexe. Douze sujets ont ~t~ explor6s. Les ~lectrodes plac~es aux points Fz, Cz, C3" et C4" ont 6t~ utilis6es pour enregistrer I'EEG avec des amplificateurs courant continu. Presser une s~rie de 6 b o u t o n s dans un ordre particulier constitue
576
la t~che motrice. I1 a ~t~ demand~ aux sujets de presser cette s~rie toutes les 20 sec aussi rapidement que possible, mais sans erreur. Des effets significatifs de temps de r~action, d'~lectrodes, et de s~quences, ainsi que les interactions ~lectrodes--s~quences ont ~t~ observes. Le BP augmente progressivement au niveau de toutes les ~lectrodes au fur et mesure que la performance s'am~liore, c'esta-dire lorsque le temps de r~ponse d & r o f t . Apr~s que la r~ponse ait atteint une asymptote, le BP enregistr~ ~ Fz, et ~ C4" diminue, tandis qu'il reste relativement constant Cz et C3". Des analyses multivari~es de covariance m o n t r e n t une relation constante entre l'am~lioration des temps de r~ponse et les modifications du BP. I would like to thank the following for their assistance and encouragement with this research; C.M. Davis, John Dickinson and specially S.G. Quapick. The author was supported by a National Research Council of Canada postgraduate scholarship while this research was conducted.
References Adams, J.A. A closed-loop theory of motor learning. J. Motor Behav., 1971, 3: 111--149. Adams, J.A. Issues for a closed-loop theory of motor learning. In: G.E. Stelmach (Ed.), Motor Control: Issues and Trends. Academic Press, New York, 1976: 87--107. Bates, J.A.V. Electrical activity of the cortex accompanying movement. J. Physiol. (Lond.), 1951, 113: 240--257. Becker, W., Iwase, K., Jurgens, R. and Kornhuber, H.H. Bereitschaftspotential preceding voluntary slow and rapid hand movements. In: W.C. McCallum and J.R. K n o t t (Eds.), The Responsive Brain. John Wright, Bristol, 1976: 81--87. Boisset, S. and Lestienne, F. The organizahon of a simple voluntary movement as analysed from its kinematics properties. Brain Res., 1974, 71: 451--457. Deecke, L., Becker, W., Grozinger, B., Scheid, P. and Kornhuber, H. Human brain potentials preceding voluntary limb movements. In: W.C. McCallum and J.R. K n o t t (Eds.), Event Related Slow Potentials of the Brain: their Relation to Behaviour. Elsevier, Amsterdam, 1973: 87--94.
M.J. TAYLOR Dincheva, H. and Harding, G.F. Changes in the readiness potential with a posterior stimulus following the motor reaction. Electroenceph. clin. Neurophysiol., 1975, 39: 671. Ford, J.M., McPherson, L. and Kopell, B.S. Differences in readiness potential associated with pushb u t t o n construction. Psychophysiology, 1973, 9: 564--567. Keele, S.W. and Summers, J.J. The structure of motor programs. In: G.E. Stelmach (Ed.), Motor Control: Issues and Trends. Academic Press, New York, 1976: 109--142. Kelso, J.A.S. and Stelmach, G.E. Central and peripheral mechanisms in motor control. In: G.E. Stelmach (Ed.), Motor Control: Issues and Trends. Academic Press, New York, 1976: 1--49. Klein, R.M. Attention and movement. In: G.E. Stelmach (Ed.), Motor Control: Issues and Trends. Academic Press, New York, 1976: 141--173. Klein, R.M. and Posner, M.J. Attention to visual and kinesthetic components of skill. Brain Res., 1974, 71: 401--411. Kornhuber, H.H., Becker, W., Taumer, R., Hoehne, O. and Iwase, K. Cerebral potentials accompanying voluntary movements in man: readiness potential and reafferent potentials. Electroenceph. clin. Neurophysiol., 1965, 26: 439. Loveless, N.E. and Sanford, A.J. Slow potential correlates of preparatory set. Biol. Psychol., 1974, 1: 303--314. McAdam, D.W. and Rubin, E.H. Readiness potential, vertex positive wave and contingent negative variation and accuracy of perception. Electroenceph. clin. Neurophysiol., 1971, 30: 511--517. McAdam, D.W. and Scales, D.M. Bereitschaftspotential enhancement with increased level of motivation. Electroenceph. clin. Neurophysiol., 1969, 27: 73--75. Schmidt, R.A. The schema as a solution to some persistent problems in motor learning theory. In: G.E. Steimach (Ed.), Motor Control: Issues and Trends. Academic Press, New York, 1976: 41--65. Sprott, D.A. Note on Evans and Anastasio on the analysis of covariance. Psychol. Bull., 1970, 73: 303--306. Teuber, H.L. The riddle of the frontal lobe function in man. In: J.M. Warren and K. Akert (Eds.), The Frontal Granular Cortex and Behaviour. McGraw-Hill, New York, 1964: 410--444. Trumbo, D. and Noble, M. Motor skill. In: B.B. Wolman (Ed.), Handbook of General Psychology. Prentice-Hall, Englewood Cliffs, N.J., 1973: 188-205. Vaughan, Jr., H.G., Costa, L.D. and Ritter, W. Topography of the human motor potential. Electroenceph, clin. Neurophysiol., 1968, 25: 1--10.
Electroencephalography and Clinical Neurophysiology , 1978, 45:568--576
© Elsevier/North-Holland Scientific Publishers, Ltd.
B E R E I T S C H A F T S P O T E N T I A L D U R I N G T H E ACQUISITION OF A S K I L L E D MOTOR TASK MARGO J. TAYLOR 1 Simon Fraser University, Burnaby, B.B. (Canada)
(Accepted for publication: March 17, 1978) Although movement-associated cortical potentials in man were first r e p o r t e d by Bates in 1951, it was n o t until 14 years later t h a t the slow negative wave preceding voluntary m o v e m e n t , the Bereitschaftspotential (BP), was first discovered and named ( K o r n h u b e r et al. 1965). Since then much research has focused on the role of the responding musculature and o f m o v e m e n t parameters on the size and cortical distribution of the BP. Research t h a t is c onc e r ned with more conative aspects of m o v e m e n t , or t hat does n o t view the BP as simply an electrophysiological correlate of the m o v e m e n t itself, is less plentiful. Becker et al. (1976) r e p o r t e d slow movements to be preceded by BPs of greater amplitude and duration than were ballistic movements. T h e y speculated t h a t m or e preparation was required for the slow, controlled m o v e m e n t s and this interpretation has been supported by o th e r researchers. Loveless and Sanford (1974) r e p o r t e d the amplitude of the wave preceding a response to be p r o p o r t i o n a l to the level of p r e p a r a t o r y set, as inferred f r o m the speed o f the reaction time. McAdam and Rubin (1971) used response accuracy as a measure o f preparation to respond and f o u n d larger BPs with accurate responses. Their subjects' post-response estimates of accuracy were positively related to the actual accuracy obtained. Ford et al. (1973) investigated BPs with qualitatively di f f er e nt b u t t o n presses. Pressing a skin-contact b u t t o n , which required more preparation than a standard push-
1 Present address: Psychology Department, McGill University, Montreal, Que., Canada.
b u t t o n , was preceded by BPs of increased amplitude and duration. McCallum ( 1 9 7 6 ) 2 report ed that as the level of task involvement increases, corresponding increases were f o u n d in the BP, with frontal areas making relatively greater contribution. A c o n f o u n d i n g variable in the study of Ford et al., however, is post-response feedback or stimulation. Their subjects received more sensory feedback concerfling the accuracy o f their response from the skinc o n t a c t b u t t o n than from the standard but t on. It has been well d o c u m e n t e d that feedback from or contingencies placed upon the m o v e m e n t affect the BP amplitude and duration (McAdam and Seales 1969; Dincheva and Hardin 1975; McCallum 1976 2; T ayl or 1976 3). Research to date with BPs has used simple repetitive movements, with little a t t e m p t to investigate the d e v e l o p m e n t of a specific m o v e m e n t or skill. Only one study has been report ed in which the voluntary response could be considered a skilled m o v e m e n t (Papakostopoulos, 1976 4). The skilled task was a single b u t t o n press, the timing of which was to be learned. Papakostopoulos f o u n d the BPs preceding correct responses were larger 2 McCallum, W.C. Relationships between the Bereitschaftspotential and the CNV. Paper presented at the 4th International Congress on ERSP, Chapel Hill, N.C., 1976. 3 Taylor, M.J. BP amplitude as a function of response contingencies. Paper presented at Society for Psychophysiological Research, San Diego, Calif., 1976. 4 Papakostopoulos, D. Electrical activity of the brain associated with skilled performance. Paper presented at the 4th International Congress on ERSP, Chapel Hill, N.C., 1976.
BP AND SKILLED TASK ACQUISITION than those preceding incorrect responses and BPs preceding the skilled task were larger than those preceding the unskilled task. Although this study offers further support for the thesis t h a t BPs reflect the subject's preparation for and accuracy in responding, problems arise with the definition of a skilled task. There was no increase in the number of correct responses over the session, y e t Papakostopoulos discussed the results in terms of developing and improving skillful performance. His subjects were n o t learning a response as much as selecting the response on cue. Research in the field of m o t o r skills, however, has shifted from an emphasis on product, or the selection of a response, to an emphasis on processes occurring during the learning of a skilled response (Trumbo and Noble 1973). For better integration of BP research with the field of m o t o r tasks, a response which can show clear improvement in performance with practice should be chosen as the voluntary movement. The m o t o r skill literature also offers much information on other aspects of a skilled task that should be considered in this type of research. The performance of a task gives rise to several sources of feedback and recently much interest has centered on the role of feedback in skilled movements. When subjects receive feedback from proprioception and vision, they are able to accurately estimate the precision of their response after the m o v e m e n t and correct errors on subsequent trials (Adams 1971, 1976; Schmidt 1976). Under conditions of ample feedback a skilled movement can rapidly approximate the response described in an instructional set; the most effective instruction to achieve this end would be one emphasizing speed (Bouisset and Lestienne 1974). As elements are added to a serial task the period required to master it lengthens and attentional requirements for the response increase. Klein and Posner (1974) report that simple discrete movements demand no attention except at initiation, but that attentional demands t h r o u g h o u t a serial m o v e m e n t
569 increase with the level of accuracy required. Improved performance is accomplished only with the allocation of additional attention (Klein 1976). The attentional demands of a repetitive serial task diminish only after the sequence is learned and response uncertainty decreases. Also, a movement t h a t a subject is prepared to make involves more attention and is performed better than an unanticipated movement. All this suggests that attention is related to preparation to respond and accuracy of the response, with obvious implications for BP research. Research in the area of m o t o r skills often parallels the research with BPs, y e t few studies have made these parallels explicit. Non-skilled self-initiated movements have been well studied by BP researchers, as have m o t o r skills by m o t o r performance, control and learning theorists. Research that can interrelate knowledge from both fields is necessary if a fuller understanding of the mechanisms involved in m o t o r skills is to evolve. This study was designed to investigate the BP and its cortical distribution during the acquisition of a skilled serial m o t o r response. Improvement in the speed of the response over trials was established as the measure of skill acquisition. A standard condition was conducted both before and after the skilled task condition to determine if any systematic changes occurred in the BP as a function of the duration of the experimental session. It was expected t h a t the BP would increase in size during learning, particularly over the frontal areas, and then diminish after acquisition. The initial rise in the BP was expected as response certainty increased; the frontal area was expected to change more rapidly to reflect preselection or the anticipatory planning of the m o v e m e n t (Teuber 1964; Kelso and Stelmach, 1976).
Method
Sub]ects Six male and 6 female paid subjects were
570
recruited from the psychology department at Simon Fraser University, ranging in age from 19 to 31 years. All were right-handed and all were naive as to the purpose of the study. Also none had had previous experience in slow potential research.
Apparatus Non-polarizable Beckman silver-silver chloride electrodes were affixed to the subject's scalp with collodion soaked gauze at Cz, Fz and two lateral placements, 5 cm lateral and 2 cm posterior to Cz (C3" and C4") (Papakostopoulos 1976). An infraorbital EOG and the EEG electrodes were referred to linked mastoids. The impedance between any two electrodes was less than 5 k~2 for all subjects. The EEG signals were amplified by Grass 7 P I A DC amplifiers, with a roll-off of 3 dB at 50 c/sec. The single trials were collected and digitized (1024 points per sweep) by a Hewlett-Packard 2116B computer. Data were collected from the ongoing record during a period 4 sec prior to and 2 sec following the initiation of the subject's response. In case of contamination by an artifact, such as eye movement, the trial was rejected and the subsequent artifact-free trial accepted in its stead.
Experimental procedure The subjects were seated in a large comfortable chair in an electrically shielded room and given a small metal box (12.5 cm × 10 cm × 8 cm) which they held on their laps. The box contained 2 rows of 3 buttons each, with 2 cm between adjacent buttons. A force of 1050 g was required to depress the buttons the necessary 5 m m for switch closure. Subject's right arm rested on the arm of the chair and was supported with a pillow to minimize involvement of the forearm when responding. All subjects used only the index finger of their right hand throughout. They were instructed to refrain from blinking and moving their eyes, particularly during the few seconds before and after a response. Subjects were asked to respond approximately every
M.J. T A Y L O R
20 sec, but were requested n o t to count or use a watch as some variability in timing was desirable. Between trials, the subject's finger rested on the first b u t t o n ensuring that the first m o v e m e n t initiating the trial was identical across trials and conditions. The first or standard condition required the subject to make a single b u t t o n press. The subjects were instructed to respond sharply as they would in a reaction time task. During this condition the subjects were given feedback, via an intercom, regarding spacing their response (i.e., whether there was either too little or too much time between trials) and eye movements (whether they were blinking or moving their eyes during trials). SubjecSs generally required very little feedback during the session. After collecting 15 trials of the standard condition instructions for the experimental condition were given. A response for this condition involved pressing a series of buttons in rapid succession. The subject was shown the pattern of b u t t o n presses, which included all 6 buttons with no repetitions (Fig. 1). The pattern was demonstrated 3 times by the experimenter but subjects were allowed no practice. Subjects were instructed to press the series of buttons as quickly as possible, always returning to the first b u t t o n upon completion. The requirement of maximizing speed w i t h o u t sacrificing accuracy was stressed. This
5
3
4
6
1
Fig. 1. The pattern of button presses constituting the serial task for the skilled task condition.
BP A N D S K I L L E D T A S K A C Q U I S I T I O N
571
emphasis led to the desired result of subjects making no errors at all in the pattern of b u t t o n presses but marked improvement over trials in the speed with which the series was pressed. Subjects were told t h a t the spacing of the responses and control of eye m o v e m e n t was to be the same as in the preceding condition, but that the duration of the experimental condition would be much longer. Forty-five trials were collected. The subjects were then informed that the standard condition was to be repeated. The earlier instructions reiterated and a final 15 trials collected. The entire experiment, including electrode application, t o o k no more than 1.5 h. Although single trial analysis would have been ideal in this study, some averaging before analyses were conducted was necessary due to the noise in single trial EEG data. In the experimental or skilled task condition, data from sets of 5 consecutive trials were averaged, yielding 9 averages for each channel for each subject. The first second of data from each trial was taken as baseline before averaging. Algebraic area measures were calculated over
3 sections of each average. The sections were defined as follows: (1) the 2 sec period prior to the response; (2) minus to plus 50 msec from the response, a peak measure; (3) the 2 sec period following response initiation. The duration of the response from the first through sixth series of b u t t o n presses and the time between the first and second b u t t o n presses were measured for all 45trials. Response time over the series of 6 presses was averaged in sets of 5 to correspond with the EEG averages for the analyses involving both measures.
Results An analysis of variance was conducted for each of the BP measures to test for difference among the two standard and the skilled task conditions. No significant condition main effect was found for the BP measure 2 sec prior to the response (F(2,20)= 1.39, NS), whereas the peak measure was significantly larger during the skilled task than during the sec
Cz
c3"
c~.
EOG
?
RESPONSE
RESPONSE
RESPONSE
RESPONSE
RESPONSE
Fig. 2. BPs averaged across subjects for each e l e c t r o d e for the 9 trial blocks of the skilled task (each wave f o r m is a c o m p o s i t e of 60 single trials; o n e trial block of 5 single trials from each of 12 subjects).
572
two standard conditions ( F ( 2 , 2 0 ) = 4 2 . 7 8 , P < 0.001). The averages across the subjects for the skilled task, for the 9 trial blocks and 4 electrodes, illustrate the general trends in the data (see Fig. 2). The influence of skill acquisition on the BP was investigated using an analysis of variance for each of the dependent area measures. Significant trial (F(8,80) = 2.34, P < 0.05) and electrode (F(3,30) = 12.64, P < 0.001) main effects and a trial by electrode interaction (F(24,240) = 1.91, P < 0.005) were found for the area measure 2 sec prior to the response and the peak area measure (respectively, F(8,80) = 2.75, P < 0.025; F(3,30) = 17.02, P <: 0.001; F(24,240) = 2.76, P<0.001). A significant electrode main effect was found for the area measure 2 sec following response initiation (F(3,30) = 8.94, P < 0.001). No significant change over the trial blocks was found for this measure ( F ( 8 , 8 0 ) = 1 . 2 2 , NS). The electrode main effects were attributable to the constant amplitude differences among the electrodes. Across subjects and trials the size of the BP increased from Fz, to C4", to C3", to C~. The greatest difference in the BP measures was between Fz and the other 3 electrodes (F(1,30) = 20.36, P < 0.001). The BP at C3" was n o t significantly larger than at C4" (F(1,30) = 3.85, P < 0.1) although the trend was in the expected direction. For the last area measure, 2 sec following response initiation, both Fz and C4" were of much smaller area than were Cz and C3" (F(1,30) = 25.86, P < 0.001). The trial main effects were due to a steady increase in the area, at all electrodes, for the first 4 averages (or 20 trials) and then some decrease, on average, over the next 4averages (or 20trials). The interaction effects were due to differences between the electrodes during the last 5 averages. At C4" and F~ the BP decreased steadily from 3 to 4 of these last 5 averages while at C~ and C3" the size dropped initially and then rebounded for 1--2 averages. At all electrodes there was an increase in the size of the BP during the last average (see Fig. 3 and Table I).
M.J. T A Y L O R -22
0 -Cz - FZ
.,,Q
•.-~..
........
O -C4""
~"
\ i
\ ....:~ \
/ ~ .....
U
...:"
,,,l
%; , -'.
:~ ......
[ ~
/ -/. :
\ ' "...
.~..
,~"
~" -
.:
? ......... ~
!,.
"a.
/""
/ - ,.
,~.. . .~
Y ." ......-"
.....
"".,~ .....
./
.,~. ......... ,,~:
..."
TRIAL
BLOCKS
Fig. 3. BP area m e a s u r e 2 sec p r i o r to t h e r e s p o n s e averaged across s u b j e c t s for t h e 9 trial b l o c k s o f t h e skilled task c o n d i t i o n .
The response time over the series of 6 b u t t o n presses decreased for all subjects over trials (F(8,80) = 28.42, P < 0.001) reaching asymptote at about the 20th trial or 4th average (Fig. 4). The response times between the first and second b u t t o n presses over the 45 trials follow the same pattern as the total response times (Fig. 5). The EEG data were also analyzed using a multivariate analysis of covariance with the
TABLE I BP area m e a s u r e s 2 sec p r i o r t o t h e r e s p o n s e for s t a n d a r d c o n d i t i o n s a n d skilled task trial blocks.
Standard Skill task Skill task Skill task Skill task Skill task Skill task Skill task Skill task Skill task Standard
I TB1 TB2 TB3 TB4 TB5 TB6 TB7 TB8 TB9 II
Fz
Cz
C3"
C4"
--6.43 --1.83 --5.08 --8.16 --10.86 --8.42 --6.50 --6.00 --5.68 --9.57 ---6.46
--12.10 --8.17 --11.26 --18.98 --20.85 --12.53 --17.51 --17.29 --11.29 --19.82 --14.32
--11.62 --8.32 --12.17 --17.40 --16.28 --12.09 --16.18 --11.76 --12.38 --14.89 --11.55
--8.05 --4.53 --11.64 --12.17 --16.47 --11.51 --9.52 --8.33 --9.62 --11.98 --9.49
BP A N D S K I L L E D T A S K A C Q U I S I T I O N
573
35"
sistent relationship between the changes in the BP and the improvement in the response times over trials {Sprott 1970).
Discussion
~,t,,,B,,,l,l,,i,,,I,l,,ss,l,,
,,*l~l,t,,,l**4,
TRIALS
5
Fig. 4. Average r e s p o n s e t i m e f r o m t h e first t h r o u g h s i x t h series o f b u t t o n presses for t h e 45 trials of t h e skilled task.
response time as the covariate. No significant trial main effects were found: F~ (F(32,281) = 1.06, NS), Cz ( F ( 3 2 , 2 8 1 ) = 1.26, NS), C3" (F(32,281) = 1.60, NS) and C4" (F(32,281) = 1.25, NS). Thus, when the EEG data across trials were adjusted for the variance contributed b y the response times the trial main effect d r o p p e d out. This demonstrates a con-
°81 ~,( •
•
~'~"
"~.~fl
"~.~'A
o•
w
TRIALS
Fig. 5. Average r e s p o n s e t i m e b e t w e e n t h e first a n d s e c o n d b u t t o n presses for t h e 45 trials o f t h e skilled task.
The results offer strong support for the hypothesis that the size and cortical distribution of the BP change with the acquisition of a skilled m o t o r task. The magnitude of the BP at all electrode locations increased steadily over the trials in which the response time was decreasing, i.e., during the acquisition of the response. The two essential requirements of the study were found for all subjects. First, a clear learning effect of the m o t o r skill and second, no significant differences between the two standard conditions, indicating that systematic changes in the BP were n o t a function of the duration of the experimental session. .As subjects acquire a skilled response, their response accuracy should improve. An increase in accuracy and preparation to respond has already been related to increased size of the BP (McAdam and Rubin 1971; Ford et al. 1973; Loveless and Sanford 1974; Papakostopoulos 1976). In all previous research, however, the measures of certainty, accuracy and preparation have been dichotomized. The subject's responses were n o t learned b u t selected and then performed either correctly or incorrectly. With the learning of a m o t o r skill gradual changes in certainty and preparation would be expected to parallel gradual improvement in performance. Subjects in the present study were fully aware of the increase in the precision of their performance over the experimental session. Thus, the systematic increases in the BP area, concurrent with the steady improvem e n t in response performance, add further credibility to the theory that the BP reflects preparation and certainty of the subject to respond. It was expected that the area of the BP 2 sec prior to the response would be larger
574 preceding the skilled response than before the single b u t t o n press. The demands placed upon the subject in terms of preparation, preselection and attention during a skilled task far outweigh those necessary for a simple movement. Also, greater amounts of information must be monitored for the performance of a skilled task compared to an unskilled task; an increase in post-response feedback has been shown to increase BP amplitude and duration (Dincheva and Harding 1975). A difference between the skilled and unskilled tasks, however, was found only for the peak measure, which does correspond to the literature which has employed only a peak amplitude measure. The explanation, though, for this lack of significance with the 2 sec BP measure may rest in the large variations in the size of the BP within the skilled task condition. For the first one or t w o initial trial blocks of the skilled task the BP was smaller at all electrode placements than it was during the standard conditions. On subsequent trials it was larger than the standard conditions except for one or t w o trial blocks after learning (see Table I). According to the hypothesis presented in this study these results would imply less accuracy and preparation for the performance of the response during the initial 5 or 10 trials of the skilled task than for the standard response. When the skilled response is acquired and uncertainty decreased, the corresponding BP is larger than in the standard conditions. The decline on some trials after learning could be attributable to decreases or lapses in the subject's attention. Klein (1976) reports attentional demands of a serial task to diminish after the task sequence has been learned. The significantly larger peak measures in the skilled task condition appeared, from visual inspection, to be concomitants of increased negativity of slope of the BP preceding skilled movement. In the standard conditions the slope of the BP was usually linear, in accordance with previous studies which employed simple movements. In the skilled task condition the slope was quadratic
M.J. TAYLOR or cubic. From this study it is impossible to determine whether the apparent difference in slope between the conditions is a function of the learning required for the skilled task or the duration of the skilled task. It was expected that the role of the frontal areas in organizing m o t o r outp.ut and in monitoring o u t p u t and feedback would be reflected in qualitatively different changes in the frontal BPs than those over the m o t o r cortex. The electrode by trial interaction, however, is due to differences only after the response had reached asymptotic levels. The frontal BPs increased at the same rate as the vertex and lateral BPs during acquisition. After response acquisition, the frontal BP, in contrast to those from Cz and C3", decreased steadily. With practice the performance of a response tends to become automatized lessening both attentional demands and the a m o u n t of feedback requisite for maintenance of the response (Keele and Summer 1976; Klein 1976). Probably the subject's involvement in the task also decreased after the goal of fast, accurate responses had been achieved, which would predict a reduction of frontal BPs (McCallum 1976). The relative size of the BPs in relation to their cortical distribution is congruent with most other research in the field. The BP was largest at Cz, larger over the m o t o r cortex than frontal areas and tended to be larger contralateral to the responding musculature. Although the peak measure corresponds to the m o t o r potential (MP) of other researchers (Vaughan et al. 1968; Deecke et al. 1973), it did n o t follow the same pattern of lateralization. As earlier research averaged a very large number of trials before the MP was evident, this difference in methodology alone could account for the lack of a lateralized peak measure in this study. The divergence between C3" and C4" after learning (see Fig. 3) suggests that the less attention a movement requires the more specific to the m o t o r projection area the BP becomes. In concurrence with this, McAdam and Rubin (1971) suggested that their finding of non-
BP AND SKILLED TASK ACQUISITION lateralized BPs was a function of the attentional demands placed u p o n their subjects. This offers an explanation of the maintenance of large BPs close to the m o t o r projection area after acquisition when BPs recorded from more remote locations decreased. Continued negativity after the initiation of the response is n o t generally reported in the literature, as a large positive wave usually follows the response within 200--300 msec. This discrepancy between the present study and the existing literature is likely due to the duration of the serial response. During the 2 sec following response initiation the response is still being performed; as can be seen in Fig. 4 the average response time, even after learning, was over 1.5 sec. In other BP research the positive wave follows the initiation of the response, b u t the responses have been simple short movements, such that it also follows the termination of the response. As the subjects acquired the skilled task and the response duration shortened to less than 2 s e c this post-BP negativity tended to decrease, although the effect was n o t significant. In this study, however, the BP reflects the level of m o t o r skill learning and, after learning, possible changes in the attention devoted to the task (Klein 1976). If one is interested in the mechanisms involved both during and after the acquisition of m o t o r skills, the BP could add a valuable dimension to the research. Potentially, it could be used to distinguish between movements that require central attentional mechanisms for initiation and performance and those that do not. The data suggest a causal relationship although, from this study, one cannot claim that the BP is an essential c o m p o n e n t of m o t o r skill acquisition, b u t certainly the question is posed for future research.
Summary Bereitschaftspotential (BP) research typically uses very simple, abrupt movements
575 that require no learning and which change little over the duration of the experimental session. The present study investigated changes in the size and cortical distribution of the BP during the acquisition of a skilled m o t o r task. Twelve subjects were employed. Electrodes at Fz, Cz, C3" and C4" were used to record the EEG with DC amplifiers. A series of 6 b u t t o n presses in a specified pattern constituted the m o t o r task. Subjects were instructed to press the series every 20 sec as quickly as possible, b u t with no errors. Significant response time, electrode and trial main effects, and electrode by trial interactions were found. The BP increased steadily at all electrodes as performance improved, i.e., as response time decreased. After the response reached asymptote the BP recorded at Fz and C4" decreased, while the BP at Cz and C3" remained relatively constant. Multivariate analyses of covariance showed a consistent relationship between the improvement in response times and the changes in the BP. This study demonstrates that the size and cortical distribution of the BP are systematically related to improved proficiency of a m o t o r response with learning.
Rdsumd Potentiels de prdparation motrice au cours de l'acquisition d'une tdche motrice complexe
Les recherches sur le potentiel de preparation motrice (BP) utilisent d'ordinaire des mouvements tr~s simples et brusques qui ne n6cessitent aucun apprentissage et qui se modifient peu sur toute la dur~e de la session exp~rimentale. Cette 6tude recherche les modifications de dimension et de distribution corticale du BP au cours de l'acquisition d'une t~che motrice complexe. Douze sujets ont ~t~ explor6s. Les ~lectrodes plac~es aux points Fz, Cz, C3" et C4" ont 6t~ utilis6es pour enregistrer I'EEG avec des amplificateurs courant continu. Presser une s~rie de 6 b o u t o n s dans un ordre particulier constitue
576
la t~che motrice. I1 a ~t~ demand~ aux sujets de presser cette s~rie toutes les 20 sec aussi rapidement que possible, mais sans erreur. Des effets significatifs de temps de r~action, d'~lectrodes, et de s~quences, ainsi que les interactions ~lectrodes--s~quences ont ~t~ observes. Le BP augmente progressivement au niveau de toutes les ~lectrodes au fur et mesure que la performance s'am~liore, c'esta-dire lorsque le temps de r~ponse d & r o f t . Apr~s que la r~ponse ait atteint une asymptote, le BP enregistr~ ~ Fz, et ~ C4" diminue, tandis qu'il reste relativement constant Cz et C3". Des analyses multivari~es de covariance m o n t r e n t une relation constante entre l'am~lioration des temps de r~ponse et les modifications du BP. I would like to thank the following for their assistance and encouragement with this research; C.M. Davis, John Dickinson and specially S.G. Quapick. The author was supported by a National Research Council of Canada postgraduate scholarship while this research was conducted.
References Adams, J.A. A closed-loop theory of motor learning. J. Motor Behav., 1971, 3: 111--149. Adams, J.A. Issues for a closed-loop theory of motor learning. In: G.E. Stelmach (Ed.), Motor Control: Issues and Trends. Academic Press, New York, 1976: 87--107. Bates, J.A.V. Electrical activity of the cortex accompanying movement. J. Physiol. (Lond.), 1951, 113: 240--257. Becker, W., Iwase, K., Jurgens, R. and Kornhuber, H.H. Bereitschaftspotential preceding voluntary slow and rapid hand movements. In: W.C. McCallum and J.R. K n o t t (Eds.), The Responsive Brain. John Wright, Bristol, 1976: 81--87. Boisset, S. and Lestienne, F. The organizahon of a simple voluntary movement as analysed from its kinematics properties. Brain Res., 1974, 71: 451--457. Deecke, L., Becker, W., Grozinger, B., Scheid, P. and Kornhuber, H. Human brain potentials preceding voluntary limb movements. In: W.C. McCallum and J.R. K n o t t (Eds.), Event Related Slow Potentials of the Brain: their Relation to Behaviour. Elsevier, Amsterdam, 1973: 87--94.
M.J. TAYLOR Dincheva, H. and Harding, G.F. Changes in the readiness potential with a posterior stimulus following the motor reaction. Electroenceph. clin. Neurophysiol., 1975, 39: 671. Ford, J.M., McPherson, L. and Kopell, B.S. Differences in readiness potential associated with pushb u t t o n construction. Psychophysiology, 1973, 9: 564--567. Keele, S.W. and Summers, J.J. The structure of motor programs. In: G.E. Stelmach (Ed.), Motor Control: Issues and Trends. Academic Press, New York, 1976: 109--142. Kelso, J.A.S. and Stelmach, G.E. Central and peripheral mechanisms in motor control. In: G.E. Stelmach (Ed.), Motor Control: Issues and Trends. Academic Press, New York, 1976: 1--49. Klein, R.M. Attention and movement. In: G.E. Stelmach (Ed.), Motor Control: Issues and Trends. Academic Press, New York, 1976: 141--173. Klein, R.M. and Posner, M.J. Attention to visual and kinesthetic components of skill. Brain Res., 1974, 71: 401--411. Kornhuber, H.H., Becker, W., Taumer, R., Hoehne, O. and Iwase, K. Cerebral potentials accompanying voluntary movements in man: readiness potential and reafferent potentials. Electroenceph. clin. Neurophysiol., 1965, 26: 439. Loveless, N.E. and Sanford, A.J. Slow potential correlates of preparatory set. Biol. Psychol., 1974, 1: 303--314. McAdam, D.W. and Rubin, E.H. Readiness potential, vertex positive wave and contingent negative variation and accuracy of perception. Electroenceph. clin. Neurophysiol., 1971, 30: 511--517. McAdam, D.W. and Scales, D.M. Bereitschaftspotential enhancement with increased level of motivation. Electroenceph. clin. Neurophysiol., 1969, 27: 73--75. Schmidt, R.A. The schema as a solution to some persistent problems in motor learning theory. In: G.E. Steimach (Ed.), Motor Control: Issues and Trends. Academic Press, New York, 1976: 41--65. Sprott, D.A. Note on Evans and Anastasio on the analysis of covariance. Psychol. Bull., 1970, 73: 303--306. Teuber, H.L. The riddle of the frontal lobe function in man. In: J.M. Warren and K. Akert (Eds.), The Frontal Granular Cortex and Behaviour. McGraw-Hill, New York, 1964: 410--444. Trumbo, D. and Noble, M. Motor skill. In: B.B. Wolman (Ed.), Handbook of General Psychology. Prentice-Hall, Englewood Cliffs, N.J., 1973: 188-205. Vaughan, Jr., H.G., Costa, L.D. and Ritter, W. Topography of the human motor potential. Electroenceph, clin. Neurophysiol., 1968, 25: 1--10.