HoukLab

The brain has a remarkable capacity for controlling movement and for thinking, utilizing both current events and memories of past experiences. Modular neuroanatomical loops link the basal ganglia and cerebellum to the cerebral cortex. There is growing evidence that both motor programs and thinking algorithms are stored in the basal ganglia and cerebellum. Practicing a specific task permits subcortical knowledge to be exported to the cerebral cortex and consolidated for more automatic processing.

My former research included: 1) microelectrode recordings of the messages transmitted along neuroanatomically defined pathways in awake behaving animals; 2) investigation of structural-functional correlations using a variety of neuroanatomical techniques; 3) use of fMRI imaging to identify the brain networks that participate in serial order recall; 4) behavioral and neuroscientific studies of on-line error correction; 5) elaboration of a theory that attempts to explain the central paradox of schizophrenia. My current work synthesizes these diverse findings and tests basic concepts by using adaptive network models of neural signal processing. Brain networks are comprised of modules, and modularity supports analogical thinking about cognitive functions.

The brain’s distributed processing modules (DPMs) have loops through the basal ganglia and through the cerebellum. The loops through basal ganglia appear to be specialized for competitive pattern classification of large input vectors from cortex. Specifically considering the loop with motor cortex, basal ganglia processing leads to the selection of a coarse action that needs to be amplified and refined by loops through the cerebellum. A combination of positive and negative feedback between motor cortex and cerebellum, together with special synaptic dynamics, provide mechanisms for the recall, initiation and execution of motor programs. My studies of the multi-level mechanisms that support these functions may lead to novel approaches for the treatment of neurological and psychiatric diseases.

The loops through the basal ganglia implement a parallel search through a very large database that has been learned from our past experiences, as reinforced by reward-related dopamine neuron input. The cerebellar cortex uses error feedback transmitted by climbing fibers to learn how to amplify and refine the basal ganglia’s ballpark approximations. I am approaching these issues from many perspectives with the goal of identifying their fundamental systems level, cellular level, and molecular level mechanisms.

My hope is to synthesize these discoveries into a coherent theory of voluntary motor control and motor learning, with extensions into cognitive neuroscience and thinking.