Questions: Cerebellum: Motor Coordination and Skill Learning

Question 1

A patient with cerebellar damage consistently overshoots when reaching for objects, landing too far rather than at the target. What does this reveal about the cerebellum's function?

• A. The cerebellum generates motor intention — without it, the patient cannot accurately represent where the target is
• B. The cerebellum amplifies muscle force signals — damage reduces the force available for controlled movement
• C. The cerebellum calibrates a predictive model of movement amplitude — without this calibration, movement extent is systematically misjudged (dysmetria)
• D. The cerebellum suppresses competing muscle groups — damage causes co-contraction and spasticity

Answer: (C)

Dysmetria — consistent over- or under-shooting — is the signature deficit of cerebellar damage. The patient knows exactly where the target is (intention intact) and can generate force (motor cortex intact), but the internal model predicting how much movement is needed is uncalibrated. Dysmetria is distinct from spasticity (option D, which is an upper motor neuron sign) and from weakness or paralysis. The key insight is that intention and execution are dissociated: cerebellar damage impairs the latter, not the former.

Question 2

In cerebellar learning, what role do climbing fibers play?

• A. They carry motor commands from the cerebral cortex to the cerebellar cortex
• B. They deliver error signals from the inferior olive that depress parallel fiber synapses when a movement prediction was wrong
• C. They carry proprioceptive feedback from muscles and joints back to the Purkinje cells
• D. They generate the efference copy used to predict upcoming sensory consequences of movement

Answer: (B)

Climbing fibers originate in the inferior olive and fire when the predicted and actual movement outcomes diverge — they carry error signals. When a climbing fiber fires alongside active parallel fiber synapses, it induces long-term depression (LTD) at those synapses, reducing their future influence on the Purkinje cell. This is the learning rule: 'these inputs predicted incorrectly — reduce their weight.' Over many repetitions, errors become rare as the forward model is refined. This is supervised learning in the brain.

Question 3

The cerebellum's use of climbing fiber error signals to modify parallel fiber synaptic weights is an example of supervised learning in the biological brain.

• T. True
• F. False

Answer: True

True — supervised learning requires a 'teacher signal' that indicates when a prediction was wrong. Climbing fibers serve exactly this role: they fire when the cerebellum's predicted movement outcome diverges from the actual outcome, and their activity selectively depresses the parallel fiber synapses that contributed to the wrong prediction. The inferior olive is the teacher; the Purkinje cells are the learning units; climbing fiber LTD is the weight update rule.

Question 4

Patients with cerebellar damage are uncertain about their intended movements and can seldom form clear motor goals.

• T. True
• F. False

Answer: False

False — cerebellar patients know exactly what they intend to do. They can describe the movement they want to make, recognize when they've missed, and try again. The deficit is entirely in execution: movements are poorly timed, misjudged in amplitude (dysmetria), or uncoordinated across multiple joints (ataxia). This dissociation between intact intention and impaired execution is one of the clearest demonstrations of what the cerebellum specifically contributes to movement.

Question 5

Why does the cerebellum need a predictive 'forward model' rather than simply relying on sensory feedback to correct movements in real time?

Model answer: Neural signals from the periphery take time to travel to the brain — proprioceptive and tactile feedback arrives tens to hundreds of milliseconds after a movement has begun. For fast, fluid movements, waiting for this feedback would produce corrective signals too late to be useful, causing jerky, poorly timed behavior. The cerebellum's forward model receives an efference copy of the motor command and uses it to predict the sensory consequences of the movement before feedback arrives. This prediction runs in parallel with actual movement, enabling smooth, rapid, automatic execution without conscious monitoring of each step.

The key insight is that the cerebellum operates predictively, not reactively. Skilled movements are fast precisely because they don't wait for error signals — the forward model preemptively adjusts the command. Cerebellar damage forces the system to rely on slow feedback, which is why cerebellar ataxia produces the characteristic decomposition of smooth movement into jerky, corrective steps.