Psychology Of Motor Skills Research Paper

From the 1960s on there was a noticeable increase in research on motor skills. Concepts of cognitive psychology turned out to be better suited for the analysis of motor behavior than, for example, concepts of behaviorism. In addition, many researchers came to entertain the idea that cognitive processes primarily serve the control of action rather than cognition per se. Finally, there was a merging of different academic disciplines such as neurobiology, physiology, neurology, biomechanics, psychology, and physical education to form what is called ‘movement science’ or ‘kinesiology.’

The psychology of motor skills can be organized in two fundamentally different ways. One organization is according to different skills such as standing, walking, speaking, and writing. The other relates to basic concepts that are applicable to real-life skills, but typically studied by means of artificial tasks that are particularly suited for their purpose. Here some major concepts will be described.

1. Autonomous And Stimulus-Based Processes Of Motor Control

There is a kind of classical controversy which, historically, came in at least two different formats. In both formats it is about the question of how critical sensory stimulation is for the production of movement. In the historically earlier format, the alternative conceptions were that of a reflex chain, according to which stimuli which arise from one elementary movement trigger the next elementary movement, and that of autonomous activity of the central nervous system in generating the basic motor patterns. In the historically later format, the reflex chain was replaced by the notion of a continuous processing of sensory feedback.

The controversy about the roles of autonomous and stimulus-based processes in motor control has dissolved. The surviving problem is how sensory information is integrated with autonomous processes. Basically, sensory information can be used continuously as a feedback signal or as a reference signal, intermittently for the updating of parameters of autonomous processes or for the triggering of new autonomous processes, and finally sensory information can be used before a movement is initiated.

There is good reason to assume that autonomous or feed-forward processes of motor control run in parallel to closed-loop processes. Such an arrangement can actually be identified anatomically in the spinal control of muscle length, and it helps in understanding some fundamental characteristics of human movements. First, closed-loop systems have a tendency to become unstable when the feedback is processed with some delay and the gain in the loop is sufficiently high. In biological systems feedback is always processed with some delay, and the high velocities of human movements would require a high gain of a closed-loop system. Nevertheless, instabilities are extremely rare. With open-loop control in parallel with closed-loop control, the closed-loop system can operate with a small gain. Hence instabilities can be avoided, but the inaccuracies of open-loop control can nevertheless be compensated.

Second, primate movements are astonishingly little affected by elimination of peripheral feedback, even by more or less complete deafferentation. Such observations suggest that peripheral feedback is not processed at all. On the other hand, the processing of peripheral feedback is clearly indicated by strong effects of distorted feedback signals. Among the best-known of such effects is stuttering, which can be induced by delayed feedback of speech signals. Such apparently contradictory findings can be predicted from a parallel arrangement of open-and closed-loop control. With this arrangement there will always be effects of distorted feedback, but the elimination of sensory feedback will affect movements only to the extent that open-loop control is inaccurate.

Sensory information is used not only as a feedback signal, but also as a reference signal which indicates desired movement characteristics. Only rarely does sensory information specify the time course of a movement in all details. More frequently, only certain characteristics, such as target positions for aiming movements, are specified, and sometimes there is essentially no sensory specification of desired movement characteristics. Consider a movement aimed at a certain target position. Before autonomous processes of motor control can be initiated, parameters related to movement characteristics such as amplitude and direction must be adjusted according to the position of the target, and other movement characteristics such as velocity must be specified according to other criteria (which often seem to be related to some kind of cost associated with performance). Such preparatory processes are often called ‘programming,’ and the autonomous processes are referred to as ‘motor programs.’ One of the continuing problems in movement research is to identify the defining characteristics of motor programs.

2. Visuo-Motor Coordination

Movements have to be adapted to the spatial layout of the environment. Transforming the visually perceived position of an object in space into appropriate motor commands which guide the hand to that particular position is a nontrivial problem. On the one hand it involves transformations between different coordinate systems, and on the other it involves transformations of kinematic movement characteristics into the appropriate torques at the joints and finally muscle activity. Typically these transformations involve ‘illposed’ problems in that there is no unique solution except when additional boundary conditions are introduced. In addition, the solutions vary across persons and across time. For example, when the task is to place the finger in a certain spatial position, the appropriate joint angles at the shoulder, elbow, and wrist depend on the lengths of the limb segments of the moving person; what exactly the required torques will be depends on the passive torques such as those which result from the action of gravity; what exactly the required innervation will be depends on the current state of the muscles, e.g., whether they are fatigued or not. Therefore, to take account of all such factors, visuo-motor coordination should exhibit a fair degree of flexibility.

Perhaps the most straightforward way to explore the flexibility of visuo-motor coordination is to modify the relation between visually perceived positions and the movements required to reach these positions. For example, spectacles with wedge prisms can be used which displace the visual world. When instructed to point (without seeing the hand) at a target which is visually displaced to the right, humans will point to the right of the target. When they are allowed to see the error at the end of each movement, it will gradually disappear—and this is a matter of minutes, not of hours. After removal of the displacing prisms, open-loop pointing exhibits a negative after-effect—the movements end to the left of the target.

The recalibration of visuo-motor coordination induced by the exposure to prismatic displacement is a fairly rapid process. Even more rapid is the recalibration of the relation between kinematic movement characteristics and the (active) torques required. For example, upward and downward movements of the hand are essentially identical kinematically, but vastly different in terms of muscle activity (and active torques). Adjustment to different weights that have to be lifted is almost instantaneous, in particular when the weight can be inferred from the appearance of the object before the movement is initiated. However, there seems to be at least one situation to which full adaptation is never achieved, namely microgravity. Microgravity changes the ‘normal’ relation between mass and weight, and here the adaptive capabilities of humans seem to fail. This may not be particularly surprising because gravitational acceleration has been a constant during human evolution, and there has never been any evolutionary pressure toward developing adaptive capabilities with respect to changes in gravitational acceleration.

Spectacles can be used to induce more severe distortions of the visual world than just displacing it by a few degrees. For example, the visual world can be rotated or inverted, or left and right can be exchanged. Adaptation to such distortions takes days rather than minutes. In spite of the full visuo-motor adaptation to severe distortions of the visual world, its appearance seems to remain different, although it is not exactly clear in what ways. The difference between visuomotor and perceptual adaptation was perhaps the first indication that sensory information for (conscious) perception of the world and motor control can be at least partially dissociated.

In the late twentieth century, both clinical and experimental evidence has been accumulated to indicate that the visual world in which we move is not exactly the world which we experience. Among the clinical evidence is the phenomenon of ‘blindsight’: patients are unable to perceive the location of a light spot in a certain part of their visual field, e.g., in the right hemifield, but nevertheless they can point to it with some degree of accuracy. Among the experimental evidence is the demonstration that induced motion apparently is ‘unknown to the hand.’ A target can be made to appear moving by way of moving a frame that surrounds it. When asked to point at the one or at the other end point of the induced motion (without seeing the hand), humans point at more or less the same position. When the induced motion is canceled by a real motion so that the target appears stationary, pointing after the target has disappeared at one or the other end point of the real motion ends at different positions. Although it is clear that ‘visual information for action’ can be separated from ‘visual information for perception,’ the precise characterization of the different visual systems involved is still a topic of research.

3. Motor Timing

Some movements have to be adapted to temporal characteristics of environmental events. An example is catching or hitting a ball. Perhaps of greatest importance is the timing of movements in playing music. These examples suggest a distinction between extrinsic and intrinsic timing. In both cases timing is of critical importance for the skill, but only in the former case it is the temporal relation with respect to some environmental event which is critical, whereas in the latter it is only the temporal relation among the movements of a sequence. Even in tasks without explicit timing requirements, temporal regularities can be observed.

A great deal of effort has been invested in identifying the visual information involved in extrinsic timing, in particular with respect to approaching objects. There is indeed visual information about time to contact with an object. In principle any ratio of a distance and its first derivative with respect to time provides an estimate of the time remaining until the distance becomes zero, provided that the first derivative is constant. Such ratios can also be computed for variables which can be sensed directly, e.g., the size of the retinal image of an object and its rate of expansion when the object approaches. This ratio in fact is involved in the extrinsic timing of, e.g., catching, but other information is likely to be involved in addition.

A kind of mixture of extrinsic and intrinsic timing can be found in the simple experimental task of tapping in synchrony with a metronome. Although this is basically an extrinsic-timing task, the timing requirements are highly regular so that intrinsic mechanisms come into play. One of the fundamental findings in synchronization tasks is the so-called negative asynchrony: the taps lead the pacing signals by about 20–50ms, depending on conditions. This asynchrony is likely to come about by different delays for processing the pacing signals and the tap feedback, so that synchronization at some central level of representation requires asynchrony at the peripheral level where the measurements are made. Negative asynchrony turns into positive synchronization errors when the frequency of the pacing signals is slowed, so that the timing of the clicks can no longer be anticipated with sufficient accuracy.

Mechanisms of intrinsic timing can be studied best in tasks without external pacing. In the simplest task, a tapping rate, which is established with the support of pacing signals, has to be continued without pacing. Among the basic results is a linear increase in variability with the mean intertap interval and a negative serial correlation between adjacent intervals, but a zero correlation between nonadjacent intervals. Hence there is a tendency for shorter and longer intertap intervals to alternate. These findings gave rise to a two-level model of motor timing which allows one to decompose the total variability of the intertap intervals into a central and a peripheral component, ascribed to a central timekeeper and motor delays, respectively. The two components are affected differentially by different task conditions and also by different kinds of brain lesions. The two-level model has been generalized to more complex unimanual and bimanual rhythms and has proven to be a useful framework for the analysis of motor timing.

For many motor skills there are no explicit timing requirements, but nevertheless there are constraints on temporal characteristics so that the timing is nonarbitrary. One of the well-researched constraints is the relation between velocity and curvature in drawing figures. A classical observation is that there is a constant velocity in drawing circles and a variable velocity in drawing ellipses; it is not only that velocity increases with the radius of curvature (as a power function with the exponent 1 3), but also that the radius of curvature changes with instructed changes in velocity. For example, when one attempts to draw a circle with high velocity in the more vertical segments and low velocity in the more horizontal segments, the result is likely to be a vertically oriented ellipse.

4. Motor Coordination

Movements are spatially and temporally adjusted not only to the environment, but also to concurrent movements of other parts of the body. Motor coordination can be in the service of the performed task. For example, in producing a certain lip aperture the positions of the upper and lower lips are correlated across a series of attempts, so that variation in the position of one lip is compensated for by the covariation of the position of the other. ‘Compensatory covariation’ seems to be a fairly robust and general mechanism for producing stable movement outcomes. Particularly intriguing is the coordination of the two hands in the service of task performance. In right-handers at least there is a clear specialization of hand use in that manipulations are assigned to the dominant (right) hand, while a supportive function is assigned to the nondominant (left) hand. The supportive function can be a simple hold function as in threading: when the needle is in the left hand, the thread will be moved toward the needle, and when the thread is in the left hand, the needle will be moved, but in each case the right hand is moved toward the left one. In more complex cases, the left hand provides something like a moving reference for the movements of the right hand, as in using a shovel or broom.

Constraints on motor coordination can also impede task performance. This is not very often noted in everyday life, except perhaps when playing a musical instrument, but it is easy to design experimental tasks which cannot be performed as instructed. For ex- ample, when concurrent movements of the two hands with different temporal and/or spatial characteristics are instructed, they tend to become too similar. This assimilation indicates the existence of cross-talk be- tween concurrent processes of motor control that is hard or impossible to overcome intentionally.

The exploration of the limits of bimanual skills has served to identify a number of constraints on coordination. The most conspicuous ones are a temporal constraint, which results in a tendency toward identical timing of concurrent left-and right-hand movements, a spatial constraint, which results in a tendency toward identical movement amplitudes, a force constraint, and a symmetry constraint. The last constraint has been studied most extensively. For example, it is an easy task to produce symmetric oscillations of the two hands. In contrast, the production of asymmetric oscillations is an easy task only when the frequency is low; with a sufficiently high frequency, asymmetric oscillations tend to switch to symmetric ones.

There are some indications that cross-talk effects in bimanual movements are asymmetric, perhaps related to the functional specialization of the two hands in bimanual actions. For example, when one hand produces a steady beat and the other hand taps as rapidly as possible or with a certain rhythm, the bimanual task is much easier when the steady beat is assigned to the left hand than when it is assigned to the right hand. Less consistently found is a smaller assimilation of the amplitudes of concurrent movements with the long amplitude being assigned to the left hand and the short amplitude to the right hand than with the opposite assignment of amplitudes. These differences are consistent with the suggestion that in bimanual tasks the left hand is specialized for long-amplitude, low-frequency movements, but the right hand for small-amplitude, high-frequency movements.

Bibliography:

1. Heuer H, Keele S W (eds.) 1996 Handbook of Perception and Action. Vol. 2: Motor Skills. Academic Press, London
2. Rosenbaum D A 1991 Human Motor Control. Academic Press, San Diego, CA
3. Schmidt R A, Lee T D 1999 Motor Control and Learning. A Behavioral Emphasis, 3rd edn. Humans Kinetics, Champaign, IL