Abstract
AbstractControl interventions (e.g., attention, inhibition) are well-known to activate frontoparietal regions and increase pupil size. Here, we characterize a different facet of cognitive-control that deactivates these very same regions and decreases pupil size. Control interventions always occur in the context of extended task episodes (e.g., ‘block of trials’). Since these extended episodes are controlled and executed as one entity, the various control interventions made across their duration are brought about via common goal-directed programs. These programs are instantiated at the beginning of the episode and are the means of organizing and controlling the numerous control interventions to be made across the task duration. These programs, therefore, are meta-control in function in that they control and organize the control interventions to be made across the task episode. Difficult episodes begin with the instantiation of more complex programs that will go on to bring about a more complex set of control interventions. Across four experiments, we show that while the instantiation of more complex control interventions during difficult task episodes expectedly activated control-related frontoparietal regions and increased pupil size, instantiation of more complex programs at the beginning of these episodes deactivated these very regions and decreased pupil size. Neural and psychophysiological signatures of the meta-control programs are thus categorically different from those of control interventions made through these programs.Significance StatementActivation of control-related frontoparietal regions and increased pupil size during difficult tasks is a well-known finding and forms the basis of our understanding of how cognition is controlled in the brain. We show that these very regions deactivate at the beginning of these very tasks that will activate them later. We suggest that this deactivation is due to meta-control structures through which control interventions across extended tasks are goal-directedly organized and instantiated. We also show that meta-control demands decrease pupil size and thus differ in their psychophysiological signatures from control demands that typically increase pupil size.
Publisher
Cold Spring Harbor Laboratory