Neurocognitive processes supporting flexible voluntary action

An overarching theme of CRC 940 relates to the operation of cognitive control processes under the antagonistic constraints of stability versus flexibility. This theme was seized during the first funding phase of project A2 by elucidating the neural and behavioral underpinnings of flexible goal-directed action under the governance of explicit instructional control processes and its interplay with the more reflexive, habit-like control of behavior. Using novel experimental approaches, two research tracks targeted two inter-related aspects of goal-directed instructional control, namely: (1) to quickly learn and modify novel behavioral rules that define prospectively by instruction how to achieve specific outcomes, that is goals and (2) to flexibly select appropriate actions to obtain variably instructed outcomes based on already learned rules (in competition with habitual action tendencies). Across these prototypical situations, we found that the instructional control of goal-directed action relies on altered functional couplings of the brain’s reward-circuit. Specifically, striatal and orbitofrontal regions were increasingly coupled with the lateral prefrontal cortex during the instruction-based learning of novel behaviors and with the angular gyrus during the instructed implementation of pre-established action-outcome contingencies. The angular gyrus in particular was most prominently relevant for creating a habit-resistant goal-directed control state. Furthermore, distinct behavioral markers were established for the various instructional control processes that can be readily used to identify process alterations in psychiatric populations, under induced stress, or following TMS application.
In the second funding phase we will extend research track 1 by implementing a novel experimental design for disentangling different within-trial BOLD activation and connectivity dynamics to uncover stimulus-related and goal-related neural processes during the earliest phase of instruction-based learning. The design will also be suited for multivariate pattern analysis to assess whether representations of instructed rules generalize to represenations of implemented rules. Research track 2, dealing with the interaction of goal-directed and habitual action tendencies, will also be a straightforward extension of the first funding phase. In particular, we will directly probe the causal role of the angular gyrus identified in the first funding phase by means of TMS. Furthermore, we will investigate the brain activation patterns underlying strategic shielding during goal-habit competition and the modulatory impact of stress on goal-habit competition and on strategic shielding during goal-habit competition. Finally, we aim to obtain a more integrated perspective on our own paradigms by systematically investigating how our behavioral measures of goal-directed control are inter-related with other conceptually closely related capacities, such as sense of agency, reliance on model-based or model-free learning, and impulsivity, as well as more general cognitive capacities such was working memory. Thereby we aim to gain a better understanding of possible factors that might drive inter-individual differences in goal-directed behavioral control.

Project Members

Principal Investigators