New fMRI studies reveal much more than the locations of brain activation during cognitive function. They indicate some of the key dynamic properties of cortical function. The fMRI results indicate that (1) the activation is modulated by the computational demand, (2) the activation is limited by resource constraints, (3) the brain activation changes with the computational demand during the time it takes to process a single cognitive problem, (4) various brain areas are synchronized with each other in their time series, and this functional connectivity is modulated by the workload (5) the activation differs with individual differences in skill or ability, and (6) additional brain areas come to be recruited as the computational demand increases or the resource supply decreases (as a result of brain damage).

The dynamics of the neural recruitment and the interactivity among the cortical components call into question the old assumption that brain function consists of a static set of decomposable systems, particularly with respect to high-level cognition. The new perspective indicates that a distributed neural system dynamically adapts itself to the moment by moment balance of computational needs and neural resources. The talk will present the beginnings of a new theory that integrates the findings within this dynamic perspective.