New learning can range from an astronaut returning from space to adjust his vestibular system to 1G, an individual associating a group of stimuli to generate a conditioned response, or someone memorizing a speech before facing an audience. In all cases, the neocortex must be engaged and signals transmitted to the cerebellar cortex to alter the synaptic weights so that the new behavior—of the vestibulo-ocular reflex, of classical conditioning, or of language acquisition—yields an automated response which is the goal of all learning. In short, how is the declarative conscious code of the neocortex converted into executable code? Sultan and Heck (2003) suggest that the mossy fibre-granular cell-parallel fibre synapses onto Purkinje neurons is such that inputs from the senses (from neocortex, brain stem, and spinal cord) can be order sequentially along a collection of parallel fibres, so that the synaptic input to a single Purkinje neuron [of which there are 15 million in human cerebellum and which contains a vast dendritic arbor (Andersen et al. 1992; Braitenberg and Atwood 1958; Nairn et al. 1989)] is synchronized to generate an optimal response whether excitatory—or inhibitory (see: Miles and Lisberger 1981). It is noteworthy that input from a single granular cell is typically insufficient to drive a Purkinje neuron, suggesting that it is the collective input from many granular elements that shapes the firing of Purkinje cells (Sultan and Heck 2003). It is the sequential timing along the parallel fibres as triggered by the mossy input that elicits new learning, which has millisecond temporal resolution (Sultan and Heck 2003). For example, when a motor command is issued by the motor cortex a signal is sent to the cerebellar mossy fibres which is then compared at a Purkinje circuit to the feedback signals from the spinal proprioceptors to assess whether the command and the feedback signal are aligned as generated via the parallel fibres (Heck and Sultan 2002). If aligned, this signals optimal performance and the end of the learning process. Sultan and Heck (2003) suggest that at least 50,000 (relatively independent) Purkinje networks throughout the cerebellar cortex of humans can be engaged simultaneously via mossy fibre input to facilitate learning (Heck and Sultan 2002; Sultan and Heck 2003). This global representation allows for all aspects of a body’s musculature to be integrated with sensory information (as conveyed from neocortex, brain stem, and spinal cord) during learning (Thach et al. 1992). That the mossy fibre input to the cerebellar cortex has global reach well beyond the circuits critical for the performance of a specific task is well established (Hasanbegović 2024), making the cerebellum an optimal learning machine to fine tune all aspects of a performance in preparation for playing a musical instrument at the highest level or for competing at the Olympic games, for instance.
- Badges
- Science topic
- Similar topics
- Methodology
- Research Methodology
- Research Papers