Abstract
Research on cognitive control commonly follows the concept of “willpower”. Accordingly, good control is considered to consist in sticking to one’s goal, and excluding every possible distraction. Yet, it is this style of processing that allows people to perform well in changing situation and adapt behavior. This has, more recently, been referred to as metacontrol which is some abstract ability that may transfer between different tasks and situations. The current study examines the neural underpinnings of this abstract ability using EEG-multivariate pattern analysis and source localization. We (i) examine whether neurophysiological activity pattern in one type of tasks predicts neurophysiological activity pattern in the other type of tasks, and (ii) whether this prediction would be more accurate for task combinations that theoretically can be assumed to share the same metacontrol bias. We show that there are indeed similarities between neurophysiological patterns and functional neuroanatomical structures across different tasks that do not share any obvious characteristics. The similarity is very specific: Metacontrol has the strongest impact during selection processes, be they related to the stimulus or the response. Moreover, similarities only show up under conditions that can be assumed to imply comparable metacontrol states. This is strong evidence for the existence of metacontrol states that are more abstract than concrete task-representations as assumed in cognitive control theories. The source localization analysis suggests that neuroanatomical structures known to form the “multiple demand (MD)” system are associated with the detected dynamics. This provides a novel view on the functional relevance of the MD.