The cognitive and neural architecture of action representations

• Applicant: Professorin Dr. Angelika Lingnau
• Subject Area: Biological Psychology and Cognitive Neuroscience
• Term: from 2018 to 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 392004592

The ability to assign meaning to actions performed by other people lies at the heart of social interactions. The aim of this proposal is to determine how we achieve this task despite huge variations in the way actions can be performed. To do so, we need to (1) examine the cognitive architecture that underlies the organization of actions and (2) the degree to which this cognitive architecture is constrained by neuronal architecture. Finally, to understand how the various different features (e.g. movement kinematics, tools, body parts) that contribute to action understanding are combined, we require (3) a systematic investigation of the functional organization of the lateral occipitotemporal cortex, a brain area known to host a variety of these action-related features. To achieve these tasks, I will transfer methods and approaches developed in the domain of object recognition and semantic cognition to the investigation of the representation of observed actions. Specifically, I will use behavioral methods (inverse multidimensional scaling, rating studies, hierarchical cluster analysis) and state-of-the-art paradigms and analysis approaches for functional magnetic resonance imaging (fMRI) data (representational similarity analysis, multivariate pattern analysis, and functional connectivity analysis). Identifying the cognitive architecture that underlies the representation of actions will provide us with a taxonomy that will be fundamental for future research in this domain. Examining the correspondence between cognitive and neural architectures will allow us to determine the computational goals of brain areas contributing to action understanding. Characterizing the representational content of key brain areas recruited during the processing of actions will provide important constraints for biologically plausible models of action recognition, and thus generate testable hypotheses for future studies.

• DFG Programme: Research Grants