Perception

Sensory rearrangement and somatosensory deafferentation experiments are being employed in this laboratory, both singly and in combination, as complementary research strategies for elucidating the role of direct sensory feedback in the learning and performance of various categories of movement.

In deafferentation experiments on adolescent monkeys, we found that movements of almost all types can be learned and performed in the absence of guidance from the periphery. Moreover, recent work with primate infants deafferented on the first day of life has demonstrated that somatosensory feedback and spinal reflexes are also not necessary, after birth, for the ontogenetic development of most types of movement performed by the forelimbs.

Sensory rearrangement studies with human subjects were directed at examining prism adaptation as a learning phenomenon. One experiment showed that, in controlling the rate of performance of an operant response, decreases in the amount of lateral displacement of vision could be used as a reward, and increases could be used as a punishment. These results support the central condition required by our previously formulated avoidance theory of prism adaptation, namely that the sensory discordance produced by prisms is aversive. In three further experiments, distribution of practice—a standard learning variable—was found to be a powerful means of manipulating such effects as (i) magnitude of adaptation, (ii) intermanual transfer, and (iii) adaptation when the subject observes his passively moved limb during the exposure period.

In an experiment combining the two techniques, the amount of prism aftereffect in monkeys with deafferented forelimbs was compared with the amount in normal monkeys as a means of evaluating the ‘proprioceptive change’ hypothesis of prism adaptation. The results provided supporting evidence for the theory, indicating that the adaptation involved a recalibration of motor—kinesthetic systems rather than a change in visual perception.

Although optical rearrangement has been extensively investigated, disarrangement has received far less attention. It is quite possible that the response of the perceptual system to the latter state, which entails a decorrelation of the normal relationship between vision and motor behavior, will lead to a better understanding of the nature of visual spatial localization.

Two studies are described in which exposure (and adaptation) to prism-displaced vision is preceded by a period of exposure to disarrangement. In the first study, concurrent exposure with a target was used. No increase in localization variability resulted from preexposure to disarrangement. However, it was found that this experience retarded the subsequent adaptation to rearrangement, relative to a condition of preexposure to normal viewing. It is argued that this result supports the existence of a cross-correlator comparator mechanism in sensorimotor coordination.

The second study was similar to the first except that there was no target-error information available during disarrangement or rearrangement. The pattern of results for the second study was different from those of the first study; this suggests that there might be important differences in the manner in which information regarding intermodality discordance and target-error is processed by the comparator.

Individual differences in adaptability to visual rearrangement have typically been ignored as a subject of investigation. However, it is clear that there is marked variability among subjects in terms of the extent to which they adapt, and possibly the form in which their adaptation takes. The present investigation examined the different ways in which subjects respond to the same experimental demands in a prism-adaptation experiment. Furthermore, a number of relevant ‘perceptual traits’ (for example the ability to fixate a visual target) were measured and found to correlate, in varying degrees, with adaptation. In general, the conclusion drawn from the latter results was that subjects who have relatively good control over their eye movements and fixations will reveal adaptation in terms mostly of the felt position of the exposed limb; subjects with relatively good control over their limb movements will adapt more in terms of their vision and hence will demonstrate a significant degree of intermanual transfer.

Changes in visually guided responses, including spatial judgments of object or limb position, which result from optical transformation of visual input are usually referred to as adaptation. The purpose of this paper is to show that the response changes observed in adaptation can be conceptualized as resulting from at least three distinct components—behavioral compensation, sensory adaptation, and visual shift. Data from a series of experiments show the nature of the interaction of behavioral compensation and sensory adaptation. Implications of this latter finding for intermanual transfer are discussed.

Arguments and evidence are presented that prism adaptation results in a third end state in addition to the ‘traditional’ components of ‘proprioceptive shift’ and ‘visual shift’. That is, under certain conditions (most importantly, ones involving error-corrective feedback), exposure to prism-displaced vision induces a motor-learning component, referred to here as an ‘assimilated corrective response’. Thus the postexposure error in target pointing, the ‘negative aftereffect’, is postulated to be the algebraic sum of proprioceptive shift, visual shift, and an assimilated corrective response—at least in certain situations. Support for the existence of this third component as a form of learning is seen in the fact that it occurs primarily when prism exposure involves target-pointing experience, and that it is apparently subject to the effects of some ‘learning variables’.

Seen through sideways-displacing prisms, the wall of a room or testing apparatus looks displaced and slanted in depth. Consequently the viewer may involuntarily treat as straight ahead a a spatial direction that is displaced from his median plane toward the optically displaced, environmentally defined axes of space. There is evidence that such cognitive shifts in the subjective straight ahead do occur and are often larger than the adaptive shifts commonly found in experiments with prisms. A straight-ahead shift would affect performance on any task that involves judgments of the straight ahead, such as judging the visual straight ahead or pointing straight ahead with eyes closed. Thus a straight-ahead shift may give the misleading impression that there is a change in visual perception, or a ‘maladaptive’ change in position sense, when there actually is none (that is, no change in performance of any other visual or proprioceptive task). A straight-ahead shift could also produce a wide variety of other unexpected, paradoxical, or misleading findings like those that have been noted in previous experiments.

It is possible to explain a number of observations of visual adaptation to optical rearrangement and other visual effects as examples of the ‘Kohnstamm phenomenon’. This is the tendency for a stressed muscle to remain innervated for a period of time after cessation of the voluntary signal to relax. When this phenomenon operates with respect to eye muscles, it may be referred to as ‘eye-muscle potentiation’. Several studies and their results are presented that demonstrate eye-muscle potentiation effects on apparent visual distance. The implications of these studies for prism adaptation are discussed.

Motor-transformation learning theory asserts that people learn through experience what stimulus transformations are under the control of their behavior. More specifically, it asserts that the parameter values of certain predetermined transformation groups are learned.

This theory was inferred, in the first place, from research on adaptation to optical rearrangement—in particular, from position-constancy adaptation in inverting-spectacles experiments, prism-displacement experiments, and in more recent computer-controlled feedback experiments. The detailed characteristics of position-constancy adaptation are found to be consistent with the theory.

Diverse consequences radiate from the theory for other human abilities, both in perception and in memory retrieval. These diverse implications are tested in studies of (a) learning to manipulate ‘objects’ in an artificial computer-controlled visual space; (b) learning to compute, in the absence of overt action, the consequences of such action; (c) learning how to access the features of prior stimuli by the execution of motor actions.

In this paper a distinction is made between stimulus-bound processes and descriptive processes. It is suggested that the processes underlying adaptation to prismatic displacement are partly stimulus bound, but that they also involve changes in centrally evoked schemata or descriptive processes. In any case, the general procedure of confronting a person with the consequences of discordant inputs can be generalized to the study of descriptive perceptual processes. A concept of the ideal perceiver, based on the mathematical idea of the group of transformations, is defined and some examples of perceptual descriptive processes or schemata are given that illustrate how the general confrontation procedure may be applied.