Differential Effects of Cerebellar Degeneration on Feedforward versus Feedback Control across Speech and Reaching Movements

Errors that result from a mismatch between predicted movement outcomes and sensory afference are used to correct ongoing movements through feedback control and to adapt feedforward control of future movements. The cerebellum has been identified as a critical part of the neural circuit underlying implicit adaptation across a wide variety of movements (reaching, gait, eye movements, and speech). The contribution of this structure to feedback control is less well understood. Although it has recently been shown in the speech domain that individuals with cerebellar degeneration produce larger online corrections for sensory perturbations than control participants, similar behavior has not been observed in other motor domains. Currently, comparisons across domains are limited by different population samples and potential ceiling effects in existing tasks. To assess the relationship between changes in feedforward and feedback control associated with cerebellar degeneration across motor domains, we evaluated adaptive (feedforward) and compensatory (feedback) responses to sensory perturbations in reaching and speech production in human participants of both sexes with cerebellar degeneration and neurobiologically healthy controls. As expected, the cerebellar group demonstrated impaired adaptation in both reaching and speech. In contrast, the groups did not differ in their compensatory response in either domain. Moreover, compensatory and adaptive responses in the cerebellar group were not correlated within or across motor domains. These results point to a general impairment in feedforward control with spared feedback control in cerebellar degeneration. However, the magnitude of feedforward impairments and potential changes in feedback-based control manifest in a domain-specific manner across individuals.SIGNIFICANCE STATEMENT The cerebellum contributes to feedforward updating of movement in response to sensory errors, but its role in feedback control is less understood. Here, we tested individuals with cerebellar degeneration (CD), using sensory perturbations to assess adaptation of feedforward control and feedback gains during reaching and speech production tasks. The results confirmed that CD leads to reduced adaption in both domains. However, feedback gains were unaffected by CD in either domain. Interestingly, measures of feedforward and feedback control were not correlated across individuals within or across motor domains. Together, these results indicate a general impairment in feedforward control with spared feedback control in CD. However, the magnitude of feedforward impairments manifests in a domain-specific manner across individuals.     

Adaptive-filter Models of the Cerebellum: Computational Analysis

Many current models of the cerebellar cortical microcircuit are equivalent to an adaptive filter using the covariance learning rule. The adaptive filter is a development of the original Marr–Albus framework that deals naturally with continuous time-varying signals, thus addressing the issue of 'timing' in cerebellar function, and it can be connected in a variety of ways to other parts of the system, consistent with the microzonal organization of cerebellar cortex. However, its computational capacities are not well understood. Here we summarise the results of recent work that has focused on two of its intrinsic properties. First, an adaptive filter seeks to decorrelate its (mossy fibre) inputs from a (climbing fibre) teaching signal. This procedure can be used both for sensory processing, e.g. removal of interference from sensory signals, and for learning accurate motor commands, by decorrelating an efference copy of those commands from a sensory signal of inaccuracy. As a model of the cerebellum the adaptive filter thus forms a natural link between events at the cellular level, such as forms of synaptic plasticity and the learning rules they embody, and intelligent behaviour at the system level. Secondly, it has been shown that the covariance learning rule enables the filter to handle input and intrinsic noise optimally. Such optimality may underlie the recently described role of the cerebellum in producing accurate smooth pursuit eye movements in the face of sensory noise. Moreover, it has the consequence of driving most input weights to very small values, consistent with experimental data that many parallel-fibre synapses are normally silent. The effectiveness of silent synapses can only be altered by LTP, so learning tasks depending on a reduction of Purkinje cell firing require the synapses to be embedded in a second, inhibitory pathway from parallel fibre to Purkinje cell. This pathway and the appropriate climbing-fibre related plasticity have been described experimentally, and its presence has implications for asymmetries and hysteresis in behavioural learning rates that are also consistent with experimental observations. These computational properties of the adaptive filter suggest that it is both powerful and realistic enough to be a suitable candidate model of the cerebellar cortical microcircuit.     

The cerebellum and neural networks for rhythmic sensorimotor synchronization in the human brain

Sensorimotor synchronization (SMS) is the rhythmic synchronization between a timed sensory stimulus and a motor response. This rather simple function requires complex cerebral processing whose basic mechanisms are far from clear. The importance of SMS is related to its hypothesized relevance in motor recovery following brain lesions. This is witnessed by the large number of studies in different disciplines addressing this issue. In the present review we will focus on the role of the cerebellum by referring to the general modeling of SMS functioning. Although at present no consensus exists on cerebellar timekeeping function it is generally accepted that cerebellar input and output flow process time information. Reviewed data are considered within the framework of the ‘sensory coordination’ hypothesis of cerebellar functioning. The idea that timing might be within the parameters that are under cerebellar control to optimize cerebral cortical functioning is advanced.     

The cerebellum in action: a simulation and robotics study

The control or prediction of the precise timing of events are central aspects of the many tasks assigned to the cerebellum. Despite much detailed knowledge of its physiology and anatomy, it remains unclear how the cerebellar circuitry can achieve such an adaptive timing function. We present a computational model pursuing this question for one extensively studied type of cerebellar-mediated learning: the classical conditioning of discrete motor responses. This model combines multiple current assumptions on the function of the cerebellar circuitry and was used to investigate whether plasticity in the cerebellar cortex alone can mediate adaptive conditioned response timing. In particular, we studied the effect of changes in the strength of the synapses formed between parallel fibres and Purkinje cells under the control of a negative feedback loop formed between inferior olive, cerebellar cortex and cerebellar deep nuclei. The learning performance of the model was evaluated at the circuit level in simulated conditioning experiments as well as at the behavioural level using a mobile robot. We demonstrate that the model supports adaptively timed responses under real-world conditions. Thus, in contrast to many other models that have focused on cerebellar-mediated conditioning, we investigated whether and how the suggested underlying mechanisms could give rise to behavioural phenomena.     

The Errors of Our Ways: Understanding Error Representations in Cerebellar-Dependent Motor Learning

The cerebellum is essential for error-driven motor learning and is strongly implicated in detecting and correcting for motor errors. Therefore, elucidating how motor errors are represented in the cerebellum is essential in understanding cerebellar function, in general, and its role in motor learning, in particular. This review examines how motor errors are encoded in the cerebellar cortex in the context of a forward internal model that generates predictions about the upcoming movement and drives learning and adaptation. In this framework, sensory prediction errors, defined as the discrepancy between the predicted consequences of motor commands and the sensory feedback, are crucial for both on-line movement control and motor learning. While many studies support the dominant view that motor errors are encoded in the complex spike discharge of Purkinje cells, others have failed to relate complex spike activity with errors. Given these limitations, we review recent findings in the monkey showing that complex spike modulation is not necessarily required for motor learning or for simple spike adaptation. Also, new results demonstrate that the simple spike discharge provides continuous error signals that both lead and lag the actual movements in time, suggesting errors are encoded as both an internal prediction of motor commands and the actual sensory feedback. These dual error representations have opposing effects on simple spike discharge, consistent with the signals needed to generate sensory prediction errors used to update a forward internal model.     

Prolonged rock climbing activity induces structural changes in cerebellum and parietal lobe

This article analyzes whether climbing, a motor activity featured by upward movements by using both feet and hands, generation of new strategies of motor control, maintenance of not stable equilibrium and adoption of long‐lasting quadrupedal posture, is able to modify specific brain areas. MRI data of 10 word‐class mountain climbers (MC) and 10 age‐matched controls, with no climbing experience were acquired. Combining region‐of‐interest analyses and voxel‐based morphometry we investigated cerebellar volumes and correlation between cerebellum and whole cerebral gray matter. In comparison to controls, world‐class MC showed significantly larger vermian lobules I‐V volumes, with no significant difference in other cerebellar vermian lobules or hemispheres. The cerebellar enlargement was associated with an enlargement of right medial posterior parietal area. The specific features of the motor climbing skills perfectly fit with the plastic anatomical changes we found. The enlargement of the vermian lobules I–V seems to be related to highly dexterous hand movements and to eye‐hand coordination in the detection of and correction of visuomotor errors. The concomitant enlargement of the parietal area is related to parallel work in predicting sensory consequences of action to make movement corrections. Motor control and sensory‐motor prediction of actions make the difference between survive or not at extreme altitude. Hum Brain Mapp 34:2707–2714, 2013. © 2012 Wiley Periodicals, Inc.     

Encoding of sensory prediction errors in the human cerebellum

A central tenet of motor neuroscience is that the cerebellum learns from sensory prediction errors. Surprisingly, neuroimaging studies have not revealed definitive signatures of error processing in the cerebellum. Furthermore, neurophysiologic studies suggest an asymmetry, such that the cerebellum may encode errors arising from unexpected sensory events, but not errors reflecting the omission of expected stimuli. We conducted an imaging study to compare the cerebellar response to these two types of errors. Participants made fast out-and-back reaching movements, aiming either for an object that delivered a force pulse if intersected or for a gap between two objects, either of which delivered a force pulse if intersected. Errors (missing the target) could therefore be signaled either through the presence or absence of a force pulse. In an initial analysis, the cerebellar BOLD response was smaller on trials with errors compared with trials without errors. However, we also observed an error-related decrease in heart rate. After correcting for variation in heart rate, increased activation during error trials was observed in the hand area of lobules V and VI. This effect was similar for the two error types. The results provide evidence for the encoding of errors resulting from either the unexpected presence or unexpected absence of sensory stimulation in the human cerebellum.     

Altered cerebellar functional connectivity mediates potential adaptive plasticity in patients with multiple sclerosis

BACKGROUND: The cerebellum is of potential interest for understanding adaptive responses in motor control in patients with multiple sclerosis because of the high intrinsic synaptic plasticity of this brain region. OBJECTIVE: To assess the relative roles of interactions between the neocortex and the cerebellum using measures of functional connectivity. METHODS: A role for altered neocortical-cerebellar functional connectivity in adaptive responses to injury from multiple sclerosis was tested using 1.5 T functional magnetic resonance imaging (fMRI) during figure writing with the dominant right hand in patients with predominantly early relapsing-remitting multiple sclerosis. RESULTS: Patients (n = 14) showed a more bihemispheric pattern of activation in motor cortex than healthy controls (n = 11). Correlations between task related signal changes in neocortical and cerebellar regions of interest were used as a measure of functional connectivity. Healthy controls showed strong functional connectivity between the left motor cortex and the right cerebellar dentate nucleus. Significant connectivity between the left primary motor cortex and the right dentate was not found in patients. However, patients had significant connectivity between the left premotor neocortex and the ipsilateral (left) cerebellar cortex (crus I), which was not found in healthy controls. CONCLUSIONS: Changes in apparent cerebellar-neocortical functional connectivity may mediate potentially adaptive changes in brain motor control in patients with multiple sclerosis. Similar changes in the cerebellum and premotor cortex have been reported in the healthy brain during motor learning, suggesting that common mechanisms may contribute to normal motor learning and motor recovery after injury from multiple sclerosis.     

Cerebrocerebellar circuits for the perceptual cancellation of eye-movement-induced retinal image motion

Despite smooth pursuit eye movements, we are unaware of resultant retinal image motion. This example of perceptual invariance is achieved by comparing retinal image slip with an internal reference signal predicting the sensory consequences of the eye movement. This prediction can be manipulated experimentally, allowing one to vary the amount of self-induced image motion for which the reference signal compensates and, accordingly, the resulting percept of motion. Here we were able to map regions in CRUS I within the lateral cerebellar hemispheres that exhibited a significant correlation between functional magnetic resonance imaging signal amplitudes and the amount of motion predicted by the reference signal. The fact that these cerebellar regions were found to be functionally coupled with the left parieto-insular cortex and the supplementary eye fields points to these cortical areas as the sites of interaction between predicted and experienced sensory events, ultimately giving rise to the perception of a stable world despite self-induced retinal motion.     

The cerebellum generates motor-to-auditory predictions: ERP lesion evidence

Forward predictions are crucial in motor action (e.g., catching a ball, or being tickled) but may also apply to sensory or cognitive processes (e.g., listening to distorted speech or to a foreign accent). According to the "internal forward model," the cerebellum generates predictions about somatosensory consequences of movements. These predictions simulate motor processes and prepare respective cortical areas for anticipated sensory input. Currently, there is very little evidence that a cerebellar forward model also applies to other sensory domains. In the current study, we address this question by examining the role of the cerebellum when auditory stimuli are anticipated as a consequence of a motor act. We applied an N100 suppression paradigm and compared the ERP in response to self-initiated with the ERP response to externally produced sounds. We hypothesized that sensory consequences of self-initiated sounds are precisely predicted and should lead to an N100 suppression compared with externally produced sounds. Moreover, if the cerebellum is involved in the generation of a motor-to-auditory forward model, patients with focal cerebellar lesions should not display an N100 suppression effect. Compared with healthy controls, patients showed a largely attenuated N100 suppression effect. The current results suggest that the cerebellum forms not only motor-to-somatosensory predictions but also motor-to-auditory predictions. This extends the cerebellar forward model to other sensory domains such as audition.     

Inhibition of the amygdala central nucleus by stimulation of cerebellar output in rats: a putative mechanism for extinction of the conditioned fear response

The amygdala and the cerebellum serve two distinctively different functions. The amygdala plays a role in the expression of emotional information, whereas the cerebellum is involved in the timing of discrete motor responses. Interaction between these two systems is the basis of the two‐stage theory of learning, according to which an encounter with a challenging event triggers fast classical conditioning of fear‐conditioned responses in the amygdala and slow conditioning of motor‐conditioned responses in the cerebellum. A third stage was hypothesised when an apparent interaction between amygdala and cerebellar associative plasticity was observed: an adaptive rate of cerebellum‐dependent motor‐conditioned responses was associated with a decrease in amygdala‐dependent fear‐conditioned responses, and was interpreted as extinction of amygdala‐related fear‐conditioned responses by the cerebellar output. To explore this hypothesis, we mimicked some components of classical eyeblink conditioning in anesthetised rats by applying an aversive periorbital pulse as an unconditioned stimulus and a train of pulses to the cerebellar output nuclei as a cerebellar neuronal‐conditioned response. The central amygdala multiple unit response to the periorbital pulse was measured with or without a preceding train to the cerebellar output nuclei. The results showed that activation of the cerebellar output nuclei prior to periorbital stimulation produced diverse patterns of inhibition of the amygdala response to the periorbital aversive stimulus, depending upon the nucleus stimulated, the laterality of the nucleus stimulated, and the stimulus interval used. These results provide a putative extinction mechanism of learned fear behavior, and could have implications for the treatment of pathologies involving abnormal fear responses by using motor training as therapy.     

Cerebellar involvement in motor but not sensory adaptation

Predictable sensorimotor perturbations can lead to cerebellum-dependent adaptation--i.e., recalibration of the relationship between sensory input and motor output. Here we asked if the cerebellum is also needed to recalibrate the relationship between two sensory modalities, vision and proprioception. We studied how people with and without cerebellar damage use visual and proprioceptive signals to estimate their hand's position when the sensory estimates disagree. Theoretically, the brain may resolve the discrepancy by recalibrating the relationship between estimates (sensory realignment). Alternatively, the misalignment may be dealt with by relying less on one sensory estimate and more on the other (a weighting strategy). To address this question, we studied subjects with cerebellar damage and healthy controls as they performed a series of tasks. The first was a prism adaptation task that involves motor adaptation to compensate for a visual perturbation and is known to require the cerebellum. As expected, people with cerebellar damage were impaired relative to controls. The same subjects then performed two experiments in which they reached to visual and proprioceptive targets while a visuoproprioceptive misalignment was gradually imposed. Surprisingly, cerebellar patients performed as well as controls when the task invoked only sensory realignment, but were impaired relative to controls when motor adaptation was also possible. Additionally, individuals with cerebellar damage were able to use a weighting strategy similarly to controls. These results demonstrate that, unlike motor adaptation, sensory realignment and weighting are not cerebellum-dependent.     

Computational Theory Underlying Acute Vestibulo-ocular Reflex Motor Learning with Cerebellar Long-Term Depression and Long-Term Potentiation

The vestibulo-ocular reflex (VOR) can be viewed as an adaptive control system that maintains compensatory eye movements during head motion. As the cerebellar flocculus is intimately involved in this adaptive motor control of the VOR, the VOR has been a popular model system for investigating cerebellar motor learning. Long-term depression (LTD) and long-term potentiation (LTP) at the parallel fiber–Purkinje cell synapses are considered to play major roles in cerebellar motor learning. A recent study using mutant mice demonstrated cerebellar motor learning with hampered LTD; the study concluded that the parallel fiber–Purkinje cell LTD is not essential. More recently, multiple forms of plasticity have been found in the cerebellum, and they are believed to contribute to cerebellar motor learning. However, it is still unclear how synaptic plasticity modifies the signal processing that underlies motor learning in the flocculus. A computational simulation suggested that the plasticity present in mossy fiber–granule cell synapses improves VOR-related sensory-motor information transferred into granule cells, whereas the plasticity in the molecular layer stores this information as a memory under guidance from climbing fiber teaching signals. Thus, motor learning and memory are thought to be induced mainly by LTD and LTP at parallel fiber–Purkinje cell synapses and by rebound potentiation at molecular interneuron–Purkinje cell synapses among the multiple forms of plasticity in the cerebellum. In this study, we focused on the LTD and LTP at parallel fiber–Purkinje cell synapses. Based on our simulation, we propose that acute VOR motor learning accomplishes by simultaneous enhancement of eye movement signals via LTP and suppression of vestibular signals via LTD to increase VOR gain (gain-up learning). To decrease VOR gain (gain-down learning), these two signals are modified in the opposite directions; namely, LTD suppresses eye movement signals, whereas LTP enhances vestibular signals.     

Making Sense of Cerebellar Contributions to Perceptual and Motor Adaptation

The cerebellum is thought to adapt movements to changes in the environment in order to update an implicit understanding of the association between our motor commands and their sensory consequences. This trial-by-trial motor recalibration in response to external perturbations is frequently impaired in people with cerebellar damage. In healthy people, adaptation to motor perturbations is also known to induce a form of sensory perceptual recalibration. For instance, hand-reaching adaptation tasks produce transient changes in the sense of hand position, and walking adaptation tasks can lead to changes in perceived leg speed. Though such motor adaptation tasks are heavily dependent on the cerebellum, it is not yet understood how the cerebellum is associated with these accompanying sensory recalibration processes. Here we asked if the cerebellum is required for the recalibration of leg-speed perception that normally occurs alongside locomotor adaptation, as well as how ataxia severity is related to sensorimotor recalibration deficits in patients with cerebellar damage. Cerebellar patients performed a speed-matching task to assess perception of leg speed before and after walking on a split-belt treadmill, which has two belts driving each leg at a different speed. Healthy participants update their perception of leg speed following split-belt walking such that the “fast” leg during adaptation feels slower afterwards, whereas cerebellar patients have significant deficits in this sensory perceptual recalibration. Furthermore, our analysis demonstrates that ataxia severity is a crucial factor for both the sensory and motor adaptation impairments that affect patients with cerebellar damage.     

The role of the cerebellum in preparing responses to predictable sensory events

Despite numerous studies on the effects of lesions of the mammalian cerebellum on coordination, adaptation and learning, the precise nature of this structure's contribution to motor control remains controversial. This paper reviews the results of a series of behavioural studies with monkeys trained to make rapid, accurate sequences of responses to visual targets. The effects of discrete cerebellar lesions on the performance of these animals is discussed in the light of recent theories about how the cerebellum might be concerned with learning to anticipate certain kinds of sensory events. Additional studies are considered that advocate sensory prediction as a fundamental cerebellar function that could contribute to many of the behavioural processes with which the cerebellum has been implicated. In particular, it is demonstrated how such information could be employed in the augmentation of motor learning by the formation of expectations about the sensory feedback arising from movements and interactions with the environment. Whilst it is argued that the cerebellum may not be unique in being able to perform such functions, comparative anatomical studies suggest that it may operate with an unequalled degree of temporal precision. Such precision forms the signature of skilled motor acts.     

Adaptive filters and internal models: Multilevel description of cerebellar function

Cerebellar function is increasingly discussed in terms of engineering schemes for motor control and signal processing that involve internal models. To address the relation between the cerebellum and internal models, we adopt the chip metaphor that has been used to represent the combination of a homogeneous cerebellar cortical microcircuit with individual microzones having unique external connections. This metaphor indicates that identifying the function of a particular cerebellar chip requires knowledge of both the general microcircuit algorithm and the chip’s individual connections.Here we use a popular candidate algorithm as embodied in the adaptive filter, which learns to decorrelate its inputs from a reference (‘teaching’, ‘error’) signal. This algorithm is computationally powerful enough to be used in a very wide variety of engineering applications. However, the crucial issue is whether the external connectivity required by such applications can be implemented biologically.We argue that some applications appear to be in principle biologically implausible: these include the Smith predictor and Kalman filter (for state estimation), and the feedback–error–learning scheme for adaptive inverse control. However, even for plausible schemes, such as forward models for noise cancellation and novelty-detection, and the recurrent architecture for adaptive inverse control, there is unlikely to be a simple mapping between microzone function and internal model structure.This initial analysis suggests that cerebellar involvement in particular behaviours is therefore unlikely to have a neat classification into categories such as ‘forward model’. It is more likely that cerebellar microzones learn a task-specific adaptive-filter operation which combines a number of signal-processing roles.     

Learning Stimulus Intervals��Adaptive Timing of Conditioned Purkinje Cell Responses

Classical conditioning of motor responses, such as the eyeblink response, is an experimental model of associative learning and of adaptive timing of movements. A conditioned blink will have its maximum amplitude near the expected onset of the unconditioned blink-eliciting stimulus and it adapts to changes in the interval between the conditioned and unconditioned stimuli. Previous studies have shown that an eyeblink conditioning protocol can make cerebellar Purkinje cells learn to pause in response to the conditioned stimulus. According to the cerebellar cortical conditioning model, this conditioned Purkinje cell response drives the overt blink. If so, the model predicts that the temporal properties of the Purkinje cell response reflect the overt behaviour. To test this prediction, in vivo recordings of Purkinje cell activity were performed in decerebrate ferrets during conditioning, using direct stimulation of cerebellar mossy and climbing fibre afferents as conditioned and unconditioned stimuli. The results show that Purkinje cells not only develop a change in responsiveness to the conditioned stimulus. They also learn a particular temporal response profile where the timing, not only of onset and maximum but also of offset, is determined by the temporal interval between the conditioned and unconditioned stimuli.     

The cerebellum as an adaptive filter: a general model?

Many functional models of the cerebellar microcircuit are based on the adaptive-filter model first proposed by Fujita. The adaptive filter has powerful signal processing capacities that are suitable for both sensory and motor tasks, and uses a simple and intuitively plausible decorrelation learning rule that offers and account of the evolution of the inferior olive. Moreover, in those cases where the input-output transformations of cerebellar microzones have been sufficiently characterised, they appear to conform to those predicted by the adaptive-filter model. However, these cases are few in number, and comparing the model with the internal operations of the microcircuit itself has not proved straightforward. Whereas some microcircuit features appear compatible with adaptive-filter function, others such as simple granular-layer processing or Purkinje cell bistability, do not. How far these seeming incompatibilities indicate additional computational roles for the cerebellar microcircuit remains to be determined.     

The role of the cerebellum in adaptation: ALE meta‐analyses on sensory feedback error

It is widely accepted that unexpected sensory consequences of self‐action engage the cerebellum. However, we currently lack consensus on where in the cerebellum, we find fine‐grained differentiation to unexpected sensory feedback. This may result from methodological diversity in task‐based human neuroimaging studies that experimentally alter the quality of self‐generated sensory feedback. We gathered existing studies that manipulated sensory feedback using a variety of methodological approaches and performed activation likelihood estimation (ALE) meta‐analyses. Only half of these studies reported cerebellar activation with considerable variation in spatial location. Consequently, ALE analyses did not reveal significantly increased likelihood of activation in the cerebellum despite the broad scientific consensus of the cerebellum's involvement. In light of the high degree of methodological variability in published studies, we tested for statistical dependence between methodological factors that varied across the published studies. Experiments that elicited an adaptive response to continuously altered sensory feedback more frequently reported activation in the cerebellum than those experiments that did not induce adaptation. These findings may explain the surprisingly low rate of significant cerebellar activation across brain imaging studies investigating unexpected sensory feedback. Furthermore, limitations of functional magnetic resonance imaging to probe the cerebellum could play a role as climbing fiber activity associated with feedback error processing may not be captured by it. We provide methodological recommendations that may guide future studies.     

Cerebellar contributions to reach adaptation and learning sensory consequences of action

When we use a novel tool, the motor commands may not produce the expected outcome. In healthy individuals, with practice the brain learns to alter the motor commands. This change depends critically on the cerebellum as damage to this structure impairs adaptation. However, it is unclear precisely what the cerebellum contributes to the process of adaptation in human motor learning. Is the cerebellum crucial for learning to associate motor commands with novel sensory consequences, called forward model, or is the cerebellum important for learning to associate sensory goals with novel motor commands, called inverse model? Here, we compared performance of cerebellar patients and healthy controls in a reaching task with a gradual perturbation schedule. This schedule allowed both groups to adapt their motor commands. Following training, we measured two kinds of behavior: in one case, people were presented with reach targets near the direction in which they had trained. The resulting generalization patterns of patients and controls were similar, suggesting comparable inverse models. In the second case, participants reached without a target and reported the location of their hand. In controls, the pattern of change in reported hand location was consistent with simulation results of a forward model that had learned to associate motor commands with new sensory consequences. In patients, this change was significantly smaller. Therefore, in our sample of patients, we observed that while adaptation of motor commands can take place despite cerebellar damage, cerebellar integrity appears critical for learning to predict visual sensory consequences of motor commands.