Joshua W. Brown

B.S., University of California, San Diego, 1996
Ph.D., Boston University, 2001
Postdoctoral Fellow, Vanderbilt University, 2001
Postdoctoral Fellow, Washington University, St. Louis, 2001-2005

Research Interests

My research focuses on the neural mechanisms of cognitive control, namely how humans monitor and flexibly direct their own behavior to achieve complex goals. I have used a variety of methodologies, including single-unit neurophysiology in awake behaving primates, studies of behavior and individual differences, functional neuroimaging, and computational neural modeling to provide unified accounts of neurophysiology, fMRI, and behavioral data. My current focus is on the neural mechanisms of error likelihood prediction and risk perception in decision-making, using combined fMRI and computational neural modeling.

Research Summary:
My interests are wide-ranging but focus on the frontal lobes. How do people and animals learn, optimize, and control goal-directed behavior in complex and changing environments? These abilities entail reinforcement learning, planning, prediction, expectation, evaluation, and sequential ordering of movements, in addition to complex sensory processing. Currently I have three main research thrusts:
1) Develop computational models of brain circuitry involved in cognitive control. My recent model of the Anterior Cingulate Cortex, or ACC (Brown & Braver, 2005, Science), suggests that ACC is critically involved in predicting the likelihood of making a mistake. Current simulations further predict that ACC activity also depends on the predicted severity of the consequences of a mistake, should one occur.
2) Test computational model predictions with fMRI. Computational modeling often provides counter-intuitive, non-trivial predictions that strongly guide empirical investigations. We are beginning to test whether ACC activity in healthy individuals reflects perceived behavioral risk, as predicted by the computational modeling work.
3) Investigate the neural bases of cognitive impairment in clinical populations using fMRI and computational modeling. We are interested in how impairments in working memory interact with possible impairments in an individual’s ability to monitor their own behavior. Computational modeling provides a framework for understanding the nature of information processing in both normal and pathological human brains.

Representative Publications

Reynolds J.R., Braver T.S., Brown J.W., and Stigchel S. (2006). Computational and neural mechanisms of task-switching. Neurocomputing, 69: 1332-1336.

Brown J.W. and Braver T.S. (2005). Learned predictions of error likelihood in the anterior cingulate cortex. Science, 307(5712): 1118-1121.

Brown J.W., Bullock D., and Grossberg S. (2004). How laminar frontal cortex and basal ganglia circuits Interact to control planned and reactive saccades. Neural Networks, 17(4): 471-510.

Ito S., Stuphorn V., Brown J.W., and Schall J.D. (2003). Performance monitoring by anterior cingulate cortex during saccade countermanding. Science, 302(5642): 120-122.

Braver T.S. and Brown J.W. (2003). Principles of pleasure prediction: specifying the neural dynamics of human reward learning. Neuron, 38(2): 150-152.

Schall J.D., Stuphorn V., and Brown J.W. (2002). Monitoring and control of action by the frontal lobes. Neuron, 36: 309-322.

Cohen J.D., Braver T.S., and Brown J.W. (2002). Computational perspectives on dopamine function in prefrontal cortex. Curr. Op. Neurobiol., 12: 223-229.