Multiple decision systems in the human brain


April 12, 2016 - 12:00pm
Northwest 243
About the Speaker
Nathaniel Daw (NYU)

The idea that the brain contains multiple distinct systems for making decisions – e.g., automatic or habitual vs. deliberative or goal-directed– is ubiquitous in behavioral and cognitive neuroscience. However, computational modeling work has until recently largely focused on a particular circuit involving dopaminergic neurons and their targets in the basal ganglia, which appears to correspond to the more automatic type of decisions as conceived by psychologists. I discuss computational modeling and experimental work which attempts to extend these models to capture the distinction between multiple decision systems, to clarify the computational mechanisms behind more deliberative control, and to permit dissociating and quantifying their contributions to trial-by-trial behavioral and neural data in human fMRI experiments. We find dissociable contributions of both mechanisms, which are differentially affected by disease, by genetics, and by laboratory manipulations. Finally, I discuss attempts to examine whether these mechanisms are implicated in psychopathology, specifically disorders of compulsion. By leveraging the efficiency of large-scale online data collection, we are able to overcome many of the issues of comorbidity that have made addressing such questions difficult.