yzed (cf. Brooks & Thach, 1981), the DN dysfunction brings about specific spatial defects, which conversely to RTs do not recover with time ( Beaubaton & Trouche, 1982). Figure 15.6 illustrates the systematic deviation in the distribution of pointing responses occurring after DN exclusion. These spatial errors in the seven monkeys indicate a shift from the control distribution which could differentiate between amplitude and directional errors. Of course the spatial impairments may be related to dynamic disturbances in the execution of movement. A relation between the hypermetric characteristics of the responses (amplitude errors) and faulty braking is a possibility in this direction. But if the range imprecision brought about the velocity impairments is accepted, one may also consider the possibility of a specific spatial impairment resulting in directional errors. The contribution of proprioceptive information to spatial calibration of the moving limb ( Paillard and Brouchon, 1974) and the role of the neocerebellum in the modulation of proprioceptive loops ( Mackay & Murphy, 1979) might be relevant points of interpretation. But the systematic misdirection of the limb could also reflect a wrong location of spatial targets due to either erroneous storage of the estimated target positions or complex disturbances in the sensorimotor coordination. In the latter case one might ask what is the role of the neocerebellum in the mechanisms for transforming spatial information into coordinated joint angular patterns ( Pellionisz & Llinas, 1980). The impairment of such a process would explain systematic errors resembling those induced by prismatic vision. Insofar as constant errors are held to be a valid index for the accuracy of central programming ( Paillard & Brouchon, 1974), the data obtained in the present experiment suggest that the cerebellum may add its own activity to other central systems responsible for motor programming, in order to ensure stability and precision ( MacKay & Murphy, 1979).
The overall data collected in the present study invite discussion of the possible participation of the neocerebellum in preparation processes. In the first place, it must be noted that the response determination which produces the association between the executive signal and the learned pointing movement continues to be performed after neocerebellar lesion. However, the existence of a delayed onset could indeed suggest a disturbance in the preparatory processes facilitating the retrieval of S-R codes. Although this hypothesis cannot be totally ruled out, the lack of interaction between the overall effect of the lesion and the number of alternative responses argues against a strong involvement of the DN in the response determination. None of the results indicate an impairment affecting that particular motor stage. It seems clear, for instance, that without the DN monkeys