Computational Properties of the Prefrontal Cortex
Nandakumar S. Narayanan, James Michael Hyman, Jeremy Keith Seamans, Erin L. Rich · Journal of Neuroscience · 2025
The term “prefrontal” appears to have been coined in the first volume of Brain in the context of describing the curious case of Phineas Gage, inspired by the skull anatomy of fish (Girard, 1851; Anon, 1878). This initial report details how prefrontal lesions defied simple explanations of disrupted movement or sensation and alluded to more complex changes in personality, motivation, and mood. Insights from this article, combined with early lesion studies in nonhuman primates, linked prefrontal function with a loss of the “faculty of attention and intelligent observation” (Ferrier, 1878). In the 150 years since these early reports, neuroscience and artificial intelligence have exploded in knowledge and technical ability. Still, the computations that endow the prefrontal cortex with its unique abilities defy easy definition. Attempts have focused on anatomical properties, such as thalamic projections, neuromodulatory input, and laminar structure, or its apparent roles in overt behavior through neuropsychological investigation. These contributions have advanced our understanding on different fronts but still have not captured the full breadth of the prefrontal cortex's role in cognition and behavior. A unifying feature of prefrontal networks is believed to be the computations it carries out—higher-order algorithms that transform immediate sensory, memory, and internal information to perform such flexible functions as inference, planning, delayed responses, … Correspondence should be addressed to Nandakumar S. Narayanan at nandakumar-narayanan{at}uiowa.edu.