Journal of Sports Sciences, September 2005; 23(9): 943 – 950
The continuous nature of timing reprogramming in an interceptive task
LUIS AUGUSTO TEIXEIRA, ELKE DOS SANTOS LIMA, & MARIANA MARI´LIA FRANZONI School of Physical Education and Sport, University of Sa˜o Paulo, Sa˜o Paulo, Brazil (Accepted 20 October 2004)
Abstract The time course of movement timing reprogramming was examined in a task requiring temporal coincidence of the conclusion of a forehand drive with the arrival of a moving luminous target at the end of an electronic trackway. The moving target departed from one end of the trackway at a constant velocity of 2 m . s71, and for a part of the trials its velocity was increased to 3 m . s71. Target velocity was modified at different moments during stimulus displacement, producing timesto-arrival after velocity increment (TAVIs) from 100 to 600 ms. The effect of specific practice on movement reprogramming was also examined. The results showed early adjustments to the action (TAVIs = 100 – 200 ms) that seemed to be stereotyped, while feedback-based corrections were implemented only at TAVIs of 300 ms or longer. Temporal accuracy was progressively increased as longer TAVIs were provided up to 600 ms. Skill training led to an overall increment of temporal accuracy, but no effect of specific practice was found. The results indicate that timing reprogramming in interceptive actions is a continuous process limited mainly by intrinsic factors: latency to initiate more effective adjustments to the action, and rate-of-movement timing reprogramming.
Keywords: Interception, movement reprogramming, moving targets, timing, visuomotor control
Introduction Unexpected changes in critical environmental events during the performance of motor tasks require the reprogramming of the ongoing motor act, with modification of its original spatiotemporal specifications. This situation is more often observed in open sport skills, in which players are induced to prepare an action plan based on environmental conditions that change rapidly, requiring a coherent modification to the movement in a short time. Demand for fast movement reprogramming is particularly evident in certain circumstances, such as unexpected deviation of the ball from its regular trajectory and velocity in interceptive tasks caused by the contact with an uneven terrain (e.g. cricket), with a net separating the two halves of a court (e.g. volleyball and racquet sports), with other players (e.g. soccer and basketball), or by the spin applied to the ball rendering its displacement different from that expected after contact with the ground (e.g. tennis). These varied examples illustrate the different circumstances in which fast movement reprogramming is required to
achieve accurate performance in unstable environments. The reprogramming of motor actions has usually been studied with aiming movements to spatial targets that abruptly change their position at some moment around the time of action onset. One aspect that has been observed in these studies is the latency for the first sign of movement adjustment. In this regard, different delays for initiation of correction have been reported, apparently reflecting distinct mechanisms of movement reorganization. Some investigations have revealed fast, supposedly unconscious movement adjustments detected as early as about 100 ms after target displacement (Brenner & Smeets, 1997; Castiello, Paulignan, & Jeannerod, 1991; Megaw, 1974; Paulignan, MacKenzie, Marteniuk, & Jeannerod, 1991). Other studies have indicated that more evident corrections to the movement take place only after a delay equivalent to a visual reaction time has elapsed [i.e. in the range of 200 – 250 ms (Barrett & Glencross, 1989; Bock & Ju¨ngling, 1999; Boulinguez & Nougier, 1999; Carnahan, Goodale, & Marteniuk, 1993; Carnahan
Correspondence: L. A. Teixeira, Escola de Educac¸a˜o Fı´sica e Esporte, Universidade de Sa˜o Paulo, Av. Prof. Mello Moraes 65, Sa˜o Paulo, S.P. 05508-900, Brazil. E-mail:
[email protected] ISSN 0264-0414 print/ISSN 1466-447X online ª 2005 Taylor & Francis Group Ltd DOI: 10.1080/02640410400023365
944
L. A. Teixeira et al.
& Marteniuk, 1994; Engel & Soechting, 2000; Georgopoulos, Kalaska, & Massey, 1981; McLeod, 1987; Soechting & Lacquaniti, 1983)], suggesting a predominant role for higher-order levels of movement control. Another important observation from research on movement reprogramming is that modification to specifications of motor parameters is a continuous rather than discrete process. Van Sonderen, Denier van der Gon and Gielen (1988) presented some evidence for this by employing the double-target (double-step) paradigm with very short inter-stimulus intervals. They found that when interstimulus intervals were 50 ms or shorter, the initial movement direction was frequently modified by the second target position already at movement onset. This modification was detected in the change of the initial movement direction: in some trials, the first portion of the movement was oriented in a direction between the primary and the secondary target. In these particular conditions of short inter-stimulus intervals, the movement was apparently initiated before the process of motor reprogramming had been concluded, resulting in an initial direction of motion somewhere in the range between the two spatial targets. These results suggest that a transition between the original and the second set of movement specifications is implemented in a continuous way, giving rise to a gradual reprogramming of the motor act. In other words, large-scale modifications to ongoing actions seem to be featured not by a complete inhibition of the original motor program and its full replacement by a new set of specifications, but by progressively adjusting movement parameters over time. From this perspective, the longer the period available for movement adjustments, the greater the chance of completing the whole process of motor reprogramming (see also Van Sonderen, Gielen, & Denier van der Gon, 1989; Van Sonderen & Denier van der Gon, 1991). The tasks usually employed to study movement reprogramming, however, have been restricted to orientation of the motion to a static spatial target, while in open sport contexts (e.g. interceptive skills) movements have to be reorganized dynamically. In these latter conditions, the displacement of a moving object is visually tracked and the individual tries to intercept it at an appropriate position in space and time that are not directly specified at the initiation of movement organization. An important difference between static and moving targets is that the former allow for a direct comparison between the position of the target and the variation of position of the aiming hand, whereas with the latter the action has to be planned as a function of the estimated time that the moving target will take to reach a certain position in the future. These distinct requirements imposed by static and moving targets may lead to different
mechanisms responsible for movement control: while a direct comparison between the hand and target positions in static settings could favour the use of feedback-based mechanisms, the indeterminacy of moving targets may entail a mode of control based predominantly on feedforward. Movement reprogramming here was studied in an interceptive task by unexpectedly changing the velocity of a moving luminous target at different moments before its arrival at a predetermined interception position. In such conditions, the difference between the original and the final target velocity induces a timing error (if no temporal adjustment is made to the motor act), which can be predicted and used as reference for detection of timing adjustments to the action. In this regard, we generated an increment in target velocity that required an augmented movement velocity in relation to the original organization of the action. If movement is appropriately reprogrammed, temporal accuracy should be similar to that observed in target motion at constant velocity, when no reprogramming is needed. In the opposite case, if movement is not reprogrammed – that is, performed in accordance with the original target velocity – actual temporal error should be similar to the difference between the arrival time for the primary and the secondary velocity. Errors in between these extreme possibilities indicate partial movement reprogramming. By modifying target velocity at different moments before target arrival, we created different periods of time for movement reprogramming that enabled us to assess the whole time course of this process. The underlying assumption in this paradigm is that if motor reprogramming is a discrete process, we would expect a delay with no change in temporal features of the action following a modification in the velocity of a moving target. Then, beyond this threshold, there should be an abrupt change in temporal accuracy, provoked by the replacement of the original motor program by a whole new set of movement specifications. If, on the other hand, movement reprogramming is a continuous process, we should observe a gradual improvement in temporal accuracy as longer periods are provided after an unexpected change in target velocity. An additional issue addressed in this study was the effect of practice on movement reprogramming. Experimental evidence has indicated that after extensive practice with full vision, individuals become able to use visual information in a more effective way (Khan & Franks, 2000; Khan, Franks, & Goodman, 1998; Teixeira, 2001; see Elliott & Lyons, 1998, for a review on the effect of specificity of learning). The results of Khan and colleagues, in particular, have suggested that one of the main benefits promoted by practice is an increased
Continuous timing reprogramming efficiency of feedback processing. This conclusion is valid for aiming movements to static targets, in which only small corrections to errors provoked by variability in movement execution are made during the action. Virtually no attention has been devoted to the effect of practice on large-scale movement corrections in response to critical environmental changes. As a consequence, there is little information on the potential to increase efficiency in the process of changing movement specifications in ongoing actions by specific learning. To address this issue, we compared a group that practised with the stmulus travelling at varying velocities (i.e. with a 50 – 50% chance of maintaining or increasing target velocity at different moments during its motion) with a group that practised with the stimulus always travelling at a constant velocity.
Methods Participants Male and female right-handed university students (n = 14) aged 19 – 27 years (mean = 20) volunteered to participate. All reported normal or corrected-tonormal vision, and were naive to the purpose of the experiment. They took part in the study after signing an informed consent form. Instrument and task The apparatus consisted of a horizontal 2-m long electronic trackway, holding a continuous series of light-emitting diodes (LEDs) 1 cm in diameter, arranged in a straight line from one end to the other, with adjacent LEDs touching each other. The LEDs were quickly turned on and off in sequence, producing a clear perception of continuous motion of a luminous spot (target) at constant velocity. Target velocity was modified by changing instantaneously the turn on/off time of the LEDs at a given moment before target arrival at the end of the trackway. The displacement characteristics of the target were controlled through a microcomputer. At the receiving end of the trackway there was a force transducer inside a tennis hemiball filled with rigid plastic material, which allowed the computer to register the time at which the sensor was touched with a precision of 1 ms. The trackway was held at approximately 70 cm above the ground, and participants stood upright beside its receiving end. The task was to hit the hemiball with a badminton racquet with a movement similar to a forward component of a tennis forehand drive at the instant the moving target arrived at the receiving end (last LED) of the trackway. The last LED of the receiving end was adjacent with the
945
hemiball to be hit. The badminton racquet was manipulated with the participant’s preferred (right) arm. Movement amplitude was set at 65 cm by asking the participants to start their movements from a vertical screen positioned behind the receiving end of the trackway at the height of the hemiball. All trials were initiated with target motion from the opposite end of the trackway with an initial velocity of 2 m × s71. This meant that under constant velocity the moving target took 1000 ms to travel from one end of the trackway to the other. Following racquet – hemiball contact, the target moved to the opposite side of the trackway at a velocity of 1 m × s71. The subsequent trial was started at the moment that the moving target reached the other end of the trackway. In some conditions, the target velocity changed instantaneously from the original 2 m × s71 to 3 m × s71, at different instants during the displacement in the participant’s direction. Such a velocity transition was easily perceived by the participants. Temporal accuracy in this task was measured as a function of the difference in time between the arrival of the target at the last LED at the receiving end of the trackway and the time at which the participant made contact with the hemiball. Experimental design and procedures Initially, the participants adjusted their position in relation to the trackway so that they could hit the hemiball with the badminton racquet oriented perpendicularly to the axis of displacement of the target. Ambient light was dim, so that the target was highly distinctive but it did not hinder vision either of the hemiball or of the arm/racquet. To become familiarized with the task, the participants performed five trials under constant target velocity and another set of five trials in which target velocity was altered at different instants before target arrival. In these trials, they also became acquainted with the way feedback was supplied, which will be described below. The experiment was conducted in three phases: pre test, practice and post test. In the pre test, participants had their temporal accuracy assessed when target velocity was kept constant and when unexpected target velocity increment occurred. For the pre test, the following conditions were employed: constant velocity of 2 m × s71, and unexpected instantaneous increment of target velocity at 150, 300, 450, 600, 750 and 900 ms before the moment of target arrival at the end of the trackway at constant velocity. Such modifications of target velocity gave rise to respective times to arrival after velocity increment (TAVIs) of 100 – 600 ms in steps of 100 ms. The differences between regular arrival times and respective TAVIs correspond to induced temporal errors by target velocity increment. Figure
946
L. A. Teixeira et al.
1 represents conditions in which target velocity was kept constant (thin line) and when target velocity was increased at 600 ms (bold line). Induced error in this circumstance corresponds to 200 ms, which is indicated in the figure by vertical dashed lines. Therefore, when velocity was increased at this instant, an optimal movement reprogramming would require a movement time 200 ms shorter than that originally planned for the initial target velocity. Table I presents a complete description of temporal periods associated with each time of target velocity increment. In one block of trials, target velocity was always constant at 2 m × s71. In the other blocks, there was a 50% chance of velocity increment, which could not be predicted by the participants. The participants were informed before each block of trials about such a possibility and the location along the trackway that the target could have its velocity increased. Increment of target velocity within a block of trials was always made at the same instant, making it easier for the participants to detect the change in the target velocity. For the condition of constant velocity the participants performed 10 trials, and for each condition of increased velocity they executed 20 trials: half under constant velocity and half under increased velocity. Therefore, there were trials under constant velocity that participants were certain about the constancy of velocity, and others in which maintenance of target velocity was uncertain. The sequence of conditions, one with constant velocity and six with velocity increments, was counterbalanced across participants. Blocks of trials were separated by rest periods of about 30 s. Immediately after the pre test, the participants practised the interceptive task in one of two conditions: target always travelling at constant velocity (‘‘constant’’ group), or 50 – 50% chance of main-
Table I. Summary of temporal landmarks across experimental conditions for constant and increased velocity (values in ms).
Time before velocity increment (ms) 850 700 550 400 250 100
Remaining Time to arrival time under after velocity increment, constant velocity (ms) TAVI (ms) Induced error 150 300 450 600 750 900
100 200 300 400 500 600
50 100 150 200 250 300
taining or increasing target velocity at different instants during target motion (‘‘increment’’ group). For the increment group, the instants of velocity increment were the same as those used in the pre test. For each of the six possible instants of velocity increment, the participants were provided with one block of 100 trials in sequence, thus accumulating 600 practice trials at the end of this phase. The sequence of blocks was the same as that used in the pre test for each participant. The constant group had the same amount of practice as the increment group in their respective condition. The computer automatically recorded the algebraic difference between the time of target arrival at the end of the trackway and the time that the hemiball was actually hit. Temporal errors were presented instantaneously to participants on the microcomputer monitor, placed at the opposite end of the trackway. Associated with visual error information, auditory feedback was also supplied through sounds emitted by the microcomputer loudspeakers indicating bandwidths (20 ms) of temporal error, which had informational and motivational purposes. Knowledge of results was also provided in the test phases to keep motivation and performance as high as possible during the assessment of movement reprogramming. The interval between blocks of trials was approximately 30 s. Three minutes after finishing the practice phase, both groups performed a post test following the same procedure as in the pre test. Dependent variables
Figure 1. Representation of one experimental condition, in which target velocity was increased at 600 ms to arrival at the interception point; after velocity transition, time to arrival (TAVI) was changed to 400 ms, inducing a delay in movement temporization of 200 ms (induced error, vertical dashed lines) if no adjustment was made.
Algebraic error and the proportional feedback-based difference between induced and observed algebraic errors (DEFb%) were analysed as dependent variables. Algebraic error indicates the direction (positive sign = delayed responses; negative sign = early responses) and magnitude of errors. In trials with increased velocity, there was an expected direction of temporal error: if less than complete amendments to the original temporal organization of
Continuous timing reprogramming the movement were made, we should observe delayed responses as compared with regular performance under constant velocity. The other variable, DEFb%, is an estimate of the proportion of movement correction in each condition. This measure is used to indicate the normalized magnitude of movement correction in relation to the perturbation (induced error) introduced in a given condition. In this sense, an algebraic error close to the induced error will indicate a value of DEFb% around zero, whereas an accurate response will achieve a value around 100. In the latter case, we interpret that a full movement reprogramming was achieved, while in the former no adjustment was made. This estimation was given by the following equation: DEFb % ¼ ½ðEi Ea þ Eu Þ=Ei 100 where Ei is the corresponding induced error for each time of velocity increment (see Table I), Ea is the average of actual algebraic error for each condition of increased velocity, and Eu is the average of algebraic error for the corresponding condition of uncertain constant velocity. This latter component in the equation was taken as an estimate of the temporal bias of the movement, assuming that performance was affected similarly by this bias both on trials with and without velocity increment. The product of the summation was normalized by the induced error to obtain a measure indicative of the proportional adjustment to the action effectively based on feedback. Results The effect of uncertainty on velocity increment on trials without alterations in velocity was assessed through a comparison of algebraic errors between the condition of certainty of constant velocity and the six conditions in which velocity could have increased but did not. This assessment was conducted using a three-way 2 (Group: constant vs. increment) 6 2 (Test: pre vs. post) 6 7 (Condition: constant velocity 6 TAVIs of 100 – 600 ms) analysis of variance (ANOVA) with repeated measures on the last two factors. This analysis indicated significant main effects for Test (F1,12 = 16.12, P 5 0.005) and Condition (F6,72 = 2.62, P 5 0.05). The main effect of Test was due to significantly earlier responses in the post test than in the pre test. Post hoc contrasts for the main effect of Condition were performed with Newman-Keuls procedures. The results indicated a single significant difference between constant velocity and possible velocity increment at 600 ms (Figure 2). In spite of this single significant difference between conditions, of note was the consistent trend of timing
947
Figure 2. Algebraic error (ms) in the pre and post test for the two groups [constant (CON) vs. increment (INC)] for a target moving at constant velocity in conditions of certainty of constant velocity (Ct) and uncertainty of velocity increment at different instants during target motion (150 – 900 ms).
errors for the direction opposite to those errors induced by velocity increment. This trend was in line with the magnitude of perturbation to timing accuracy, as will be shown below. This is indicative of an overall bias in movement organization that was related to the instant of possible velocity increment. Similarity of performance between groups suggests that both conditions of practice led to an equivalent strategy of temporal organization of the original motor program. Algebraic errors in conditions of changed velocity were analysed using a three-way 2 (Group) 6 2 (Test) 6 7 (Condition: constant velocity 6 TAVIs of 100 – 600 ms) ANOVA with repeated measures on the last two factors. The results indicated significant main effects for Test (F1,12 = 19.16, P 5 0.001) and Condition (F6,72 = 26.95, P 5 0.0001), in addition to a significant three-way interaction (F6,72 = 3.98, P 5 0.005). The main effect of Test was due to a systematic reduction of error for both groups from the pre to the post test. Post hoc contrasts for the main effect of Condition indicated that performance with target in constant velocity was characterized by significantly lower errors compared with all conditions of increased velocity. Performance across these conditions produced an inverted U-shaped curve, with significantly larger errors at TAVIs of 200 – 300 ms than in the other conditions. Except for TAVIs of 200 – 300 ms and 500 – 600 ms, all other contrasts between adjacent conditions were significant. The three-way interaction was due to a single significant between-groups difference in TAVI of 300 ms in the pre test (Figure 3). The absence of an interaction between the main factors Group and Condition revealed that the two conditions of practice led to a similar capacity to reprogram the timing component of the action. Analysis of DEFb% was performed only in the post test and was conducted using a two-way 2 (Group) 6 6 (Condition: TAVIs of 100 – 600 ms) ANOVA
948
L. A. Teixeira et al.
with repeated measures on the second factor. The results indicated only a significant main effect for Condition (F5,60 = 24.77, P 5 0.0001). Post hoc contrasts revealed that all comparisons were significant except in the first (TAVI = 100 – 200 ms) and the last two (TAVI = 400 – 500 ms and 500 – 600 ms) adjacent conditions. As can be seen in Figure 4, up to a TAVI of 200 ms, there was a constant rate of movement correction, whereas in the following periods a progressive increment in the proportion of correction was observed, reaching a value close to 100% for the TAVI of 600 ms. Discussion The results reveal different aspects relevant for understanding movement reprogramming. For latency of timing adjustments in particular, we identified two distinct periods. The first corresponds to small corrections that can be implemented in short periods of time, and the second is characterized by larger corrections that need longer times to be put in action. As observed at TAVIs of 100 – 200 ms, the
Figure 3. Algebraic error (ms) for the constant (CON) and increment (INC) groups, in the pre and post test, as a function of TAVI (100 – 600 ms); constant velocity (CV) is presented as a reference for the regular unperturbed performance.
participants were able to adjust the timing component of their actions at approximately 10 – 30% of the complete correction required to reach the performance observed under certain constant target velocity. A point worth noting was the constant proportion of timing reprogramming between TAVIs of 100 and 200 ms. The absence of gain in temporal accuracy with a longer time for movement correction was peculiar to this short period, and is indicative of stereotyped corrections that are not susceptible to graded adjustments based on visual feedback. Apparently, visual information up to 200 ms before target arrival was able to automatically trigger a fast response (i.e. an increment of movement velocity), but had only a very limited capacity to integrate visual afference and movement control. This finding is in line with previous results showing fast adjustments to abrupt changes in target position (Brenner & Smeets, 1997; Castiello et al., 1991; Megaw, 1974; Paulignan et al., 1991). Analysis of TAVIs in the range of 300 – 600 ms showed a distinct profile regarding the early adjustments observed in the period of 100 – 200 ms. The progressive improvement in accuracy across longer TAVIs indicates that the reprogramming of movement timing specification occurred as a continuous process, which was apparently concluded in approximately 600 ms. One of the main features in the data that leads to this conclusion was the gradual approach to accurate timing performance, without modification of the direction of timing errors during this process. On the other hand, if response reprogramming were made in a discrete way, we should see an abrupt variation of timing accuracy following the structural delay for a voluntary change of movement specifications. Rather, we observed a progressive reduction in the magnitude of delayed responses as TAVIs increased. As revealed in the analysis, a main period for response reprogramming occurred between TAVIs of 200 and 400 ms, when
Figure 4. Proportional delta error for feedback-based corrections (DEFb%) for each experimental group in the post test as a function of TAVI (100 – 600 ms).
Continuous timing reprogramming timing accuracy was improved at a higher rate. In this period, the participants had adjusted their response by approximately 70% in relation to the error induced by the original target velocity. After this period, we found a less steep slope of response adjustment, leading to an additional period of 200 ms for reacquisition of performance similar to that seen in the condition of certain constant velocity. The different profiles of the corrections observed between TAVIs of 100 – 200 ms and TAVIs of 300 – 600 ms seem to have captured the distinction between fast triggered and graded feedback-based corrections that have been identified in previous studies of aiming movements (cf. Barrett & Glencross, 1989; Boulinguez & Nougier, 1999; Carnahan et al., 1993; Carnahan & Marteniuk, 1994; Paulignan et al., 1991). A gradual reacquisition of temporal accuracy after velocity change could be explained by the time required for observation of target motion after velocity transition to estimate the new time to arrival. When target velocity was changed earlier, longer periods were provided not only for a motor process of reprogramming the action, but also for observing the target motion and extracting perceptual information. Carnahan and colleagues (Carnahan & McFadyen, 1996; Mason & Carnahan, 1999) have shown that in a task of grasping a moving object under different displacement velocities, when distance of object displacement was controlled object velocity affected all transport-dependent kinematic measures of manual movements. In contrast, when viewing time was controlled (Mason & Carnahan, 1999), movement kinematics was no longer affected by object velocity. These findings are suggestive of the role played by viewing time in the organization of interceptive movements, which might represent a rate limiter for precise reprogramming of the action. Other studies, however, have indicated that movement reprogramming occurs continuously also in tasks holding discrete transitions between external specifications for the action (Van Galen & Weber, 1998; Hening, Favilla, & Ghez, 1988; Van Sonderen et al., 1988, 1989; Van Sonderen and Denier van der Gon, 1991). In these studies, the main perceptual requirement was to detect the presence of the new target, in which perceptual analysis of target features was a simple discrete component in movement reorganization. Even with such low perceptual requirements, the results indicated a gradual reacquisition of accurate performance as a function of time after target change. The comparison between the times for the completion of movement reprogramming in the present study and those achieved in earlier investigations with abrupt target displacement suggests that the long delays observed here are the result of a combination of the effects of the perceptual and the motor components.
949
Van Sonderen and Denier van der Gon (1991), for example, estimated that the minimum time for correctly reorienting movement initiation was in the order of 250 ms, while Hening et al. (1988) estimated times in the range of 250 – 400 ms for a complete response specification. Our results showed that at a TAVI of 400 ms, most response reprogramming had been accomplished. Nevertheless, movement reprogramming was on course up to 600 ms after target velocity increment. Such a discrepancy in the time to fully reprogram a motor act may be related to the perceptual analysis required for a precise estimation of the time to arrival after target velocity increment. In this case, the viewing time required for perceptual analysis would increase the total time needed to implement appropriate new timing specifications. Yet, as discussed at the outset, one important difference between static and moving targets is that the former allow for a direct comparison between the position of the static target and the position of the aiming hand, whereas with the latter the action has to be planned as a function of the anticipated time that the moving target takes to reach a given position. Based on this rationale, the use of perceptual anticipation in interceptive tasks may also require extra attentional resources, adding to the period required for reprogramming actions directed to static targets. Specifically for graded corrections, it is interesting that in all conditions with changed target velocity algebraic error was significantly higher compared with the condition in which the target moved at certain constant velocity. This suggests that the original movement program was prepared from the start of the target motion, even with a relatively high probability of velocity change. As the participants knew a priori the possible time of velocity increment, they appear to have been unable to prevent the preparation of the action up to the critical instant of velocity transition. This strategy of retarding movement organization would avoid the need for motor reprogramming, and might improve temporal accuracy. In this sense, visual information of the target seems to have elicited the preparation of the action with a visual exposition as short as 100 ms, as observed with the TAVI of 600 ms. Despite different requirements of movement reprogramming during the practice phase, both groups presented very similar performance profiles throughout all conditions of target velocity increment. This similarity of performance between groups showed that practice with velocity alteration did not lead to an increased capacity either to make early movement corrections or to implement more accurate motor adjustments. This finding reveals that timing reprogramming was not rendered more efficient by exercising this sensorimotor function
950
L. A. Teixeira et al.
during practice. Yet, the similarity of performance between the constant and the increment groups under uncertain constant velocity conditions reveals that both experimental groups employed a similar strategy of movement control, which in general biased performance towards early responses. Therefore, one of the main conclusions we can draw from these results is that different conditions of practice, requiring movement timing reprogramming or not, had a similar impact on the capacity to make movement corrections. In conclusion, practising the transition between timing specifications did not make this visuomotor function more efficient, which suggests that reprogramming of the interceptive action studied here was limited essentially by intrinsic factors: latency to start more effective graded adjustments to the action, and rate of movement timing reprogramming. The time required for the complete movement timing reprogramming has important implications for interceptive skills in general, since unanticipated changes in critical environmental events are expected to be dealt with through a slow process of modification of the original movement specifications. If this interpretation is correct, with unexpected changes in the time to arrival of a moving target (like a ball in many open sport skills) at the expected interception position, the movement performed will be the result of an incomplete process of timing reorganization. As a consequence, the shorter the time for movement reprogramming, the more marked the failure in movement synchronization.
Acknowledgements This study was supported by FAPESP, Brazil, through a grant (#11141-0) provided to the first author, and by providing studentships to the second (#08750-1) and third (#13668-1) authors. We are thankful to Dr Ronald Ranvaud for his detailed comments on an earlier draft of this article.
References Barrett, N. C., & Glencross, D. J. (1989). Response amendments during manual aiming movements to double-step targets. Acta Psychologica, 70, 205 – 217. Bock, O., & Ju¨ngling, S. (1999). Reprogramming of grip aperture in a double-step virtual grasping paradigm. Experimental Brain Research, 125, 61 – 66. Boulinguez, P., & Nougier, V. (1999). Control of goal-directed movements: The contribution of orienting of visual attention and motor preparation. Acta Psychologica, 103, 21 – 45. Brenner, E., & Smeets, J. B. J. (1997). Fast responses of the human hand to changes in target position. Journal of Motor Behavior, 29, 297 – 310.
Carnahan, H., Goodale, M. A., & Marteniuk, R. G. (1993). Grasping versus pointing and the differential use of visual feedback. Human Movement Science, 12, 219 – 234. Carnahan, H., & Marteniuk, R. G. (1994). Hand, eye, and head coordination while pointing to perturbed targets. Journal of Motor Behavior, 26, 135 – 146. Carnahan, H., & McFadyen, B. J. (1996). Visuomotor control when reaching toward and grasping moving targets. Acta Psychologica, 92, 17 – 32. Castiello, U., Paulignan, Y., & Jeannerod, M. (1991). Temporal dissociation of motor responses and subjective awareness. Brain, 114, 2639 – 2655. Elliott, D., & Lyons, J. (1998). Optimizing the use of vision in skill acquisition. In J. Piek (Ed.), Motor control and human skill: A multidisciplinary perspective (pp. 57 – 72). Champaign, IL: Human Kinetics. Engel, K. C., & Soechting, J. F. (2000). Manual tracking in two dimensions. Journal of Neurophysiology, 83, 3483 – 3496. Georgopoulos, A. P., Kalaska, J. F., & Massey, J. T. (1981). Spatial trajectories and reaction times of aimed movements: Effects of practice, uncertainty, and change in target location. Journal of Neurophysiology, 46, 725 – 743. Hening, W., Favilla, M., & Ghez, C. (1988). Trajectory control in targeted force impulses. V. Gradual specifications of response amplitudes. Experimental Brain Research, 71, 16 – 128. Khan, M. A., & Franks, I. M. (2000). The effect of practice on component submovements is dependent on the availability of visual feedback. Journal of Motor Behavior, 32, 227 – 240. Khan, M. A., Franks, I. M., & Goodman, D. (1998). The effect of practice on the control of rapid aiming movements: Evidence for an interdependency between programming and feedback processing. Quarterly Journal of Experimental Psychology, 51A, 425 – 444. Mason, A. H., & Carnahan, H. (1999). Target viewing time and velocity effects on prehension. Experimental Brain Research, 127, 83 – 94. McLeod, P. (1987). Visual reaction time and high-speed ball games. Perception, 16, 49 – 59. Megaw, E. D. (1974). Possible modification to a rapid ongoing programmed manual response. Brain Research, 71, 425 – 441. Paulignan, Y., MacKenzie, C., Marteniuk, R., & Jeannerod, M. (1991). Selective perturbation of visual input during prehension movements. 1. The effects of changing object position. Experimental Brain Research, 83, 502 – 512. Soechting, J. F., & Lacquaniti, F. (1983). Modification of trajectory of a pointing movement in response to a change in target location. Journal of Neurophysiology, 49, 548 – 564. Teixeira, L. A. (2001). Integrac¸a˜o visomotora no controle de tarefas sincronizato´rias [Visuomotor integration in the control of synchronization tasks]. In L. A. Teixeira (Ed.), Avanc¸os em Comportamento Motor (pp. 201 – 224). Rio Claro: Movimento, Brazil. Van Galen, G. P. V., & Weber, J. F. (1998). On-line size control in handwriting demonstrates the continuous nature of motor programs. Acta Psychologica, 100, 195 – 216. Van Sonderen, J. F., & Denier van der Gon, J. J. (1991). Reactiontime-dependent differences in the initial movement direction of fast goal-directed arm movements. Human Movement Science, 10, 713 – 726. Van Sonderen, J. F. van, Denier van der Gon, J. J., & Gielen, C. C. A. M. (1988). Conditions determining early modification of motor programmes in response to changes in target location. Experimental Brain Research, 71, 320 – 328. Van Sonderen, J. F., Gielen, C. C. A. M., & Denier van der Gon, J. J. (1989). Motor programmes for goal-directed movements are continuously adjusted according to changes in target location. Experimental Brain Research, 78, 139 – 146.
