Central feedback pathways for motor control were studied by recording EMG responses to sudden upper limb displacements in humans and monkeys using a precision torque motor to generate step load changes. Normal human subjects showed three short-latency EMG responses (Ml, M2 and M3) which appear to correspond to those recorded from trained monkeys. The M2 and M3 components, thought to represent feedback in supraspinal pathways, were significantly increased when the subjects were instructed to actively compensate for the load changes.

Parkinsonian patients with rigidity showed evidence of markedly increased feedback over the interval for the M2 and M3 responses and appeared to have lost the ability to modulate feedback according to the motor task being performed. The results are discussed with reference to recent research on motor control mechanisms in primates and a tentative model for the basis of Parkinsonian rigidity is proposed.

Experiments in humans and in monkeys have indicated that load perturbations, occurring during voluntary movements and postural activity, may be automatically compensated for. Overall muscle stiffness opposing load changes is determined by the viscoelastic properties of the muscle, by segmental reflex actions and finally by long-loop reflexes. Under certain circumstances, for instance when the subject or the experimental monkey is “prepared” to counteract perturbations which are unpredictable in time, the long-loop “reflexes” appear to be responsible for most of the corrective muscle tension. Experiments in anaesthetized monkeys revealed that signals from stretch afferents reach neurons of the motor cortex, possibly via a relay in the cortical area 3a. The latencies of these responses to well controlled muscle stretches were in the same range as motor cortical cell discharges recorded in alert monkeys subjected to load perturbations. Furthermore, these responses of cells in the motor cortex also had the appropriate timing to indicate a causal relationship with the long-latency electromyographic responses to load changes referred to above. These experimental results therefore strongly support the hypothesis, first proposed by Phillips (1969), of a transcortical servoloop adjusting motor cortical output according to the load conditions in which movements are performed.

The major advantage of transcortical regulations as opposed to segmental regulations, seems to be a powerful gain control acting at the cortical level; it was repeatedly shown that the long-loop reflexes are strongly modifiable and under voluntary control. It is suggested that an adaptive gain control at the cortical level is a prerequisite to preserve the complex capabilities of the motor cortex as the chief "executive" for skilled, preprogrammed movements. A loss of this adaptive gain control may be, at least partly, the cause of motor disorders such as rigidity in Parkinsonian patients, as reported by Tatton and Lee (1975). It is suggested that further investigations of the control of transcortical reflexes may aid in the understanding of the pathophysiology of motor disabilities.

Decorticate spasticity in the squirrel monkey was chosen as a convenient laboratory model of spasticity capable of quantitative assessment upon which to evaluate various currently popular clinical spasmolytic measures. The effects of a wide variety of cortical lesions were studied involving primary and supplementary motor, premotor and parietal cortex unilaterally and bilaterally, measuring muscle tone with the evoked integrated E.M.G. technique. Measurable spasticity resulted only if primary motor cortex was ablated bilaterally usually but not always preferentially involving biceps brachii and quadriceps. Resulting postures were variable offering no justification for the term “decorticate posture”. The integrated evoked E.M.G. was proportional to rate of stretch and chiefly phasic in type as in hemiplegic man. Stereotactic dentatectomy resulted in profound ipsilateral reduction in this spasticity, but was without effect in intercollicular or anemic decerebrate cats. The mechanism of the spasticity and of the cerebellar effects are discussed.

It was with considerable hesitation that I finally accepted the kind invitation of Vernon Brooks and the Organizing Committee of this Symposium to say a few words at the end giving some of my impressions of the presentations and discussion and reflections on the remarkable advances in our knowledge of motor systems that has occurred during recent years. I have been asked also if I would try to place some of these findings from experimental laboratories in the context of clinical experience with diseases and disorders of motor control. I would like to draw particular attention to the importance of state dependent reactions in which patterns of behavior are set up, even including spinal reflexes, by present programs depending on the general or directional reactive state of the organism.