to represent positions within the bounds of the reference system, and three targets were selected to represent positions outside the bounds of the reference system. Providing the Stelmach and Larish ( 1980) interpretation is correct, the expected pattern of results is as follows: same- and switched-limb accuracy will be equivalent when limb orientation remains inside the egocentric reference system, whereas same-limb accuracy will be superior when limb orientation is beyond the influence of the reference system. In addition, switched-limb accuracy inside the reference system will be superior to that outside.
At locations defined a priori as inside egocentric space (location near the body midline), accuracy was not different between the limb conditions. At locations defined a priori as outside egocentric space, the same-limb condition was superior to the switched-limb condition. On the basis of these data, one is compelled to believe that efficient motor control by the spatial code is possible only when localization is made within the egocentric reference system. These data support MacNeilage's target hypothesis but suggest restrictions to its generality. When body referent points are available, the accuracy of spatial positioning proceeds independently of the movement originally responsible for creating the spatial location code. Moreover, beyond the limits of this body-based reference system, accurate spatial positioning becomes dependent on direct proprioceptive information.
In general, the reviewed studies imply that accurate limb orientation need not be executed on the basis of remembered sensory consequences and suggest that a spatial coordinate system may mediate the storage of movement information. From the perspective of economy, such an assertion is appealing, as it permits a certain degree of flexibility in the central nervous system's ability to represent movement information.
One of the major contemporary issues in motor-behavior research concerns the development of an internal memory representation, postulated as necessary for guiding and controlling movement. This chapter addressed this issue by examining the vast amount of research on the short-term retention of simple motor responses. In contrast to the previous chapter the discussion was not limited to any one theory but presented several viewpoints attempting to stress important findings. From a broad perspective it was pointed out that much of the work in this area focused on developing a memory representation of movement information that was similar in theoretical context to that of Adams' view for developing a memory reference. Most of this work was concerned with examining the causes of forgetting and stressed strategies to improve memory representation, including the importance of cognitive activity.