Emotions are often assumed to manifest in subcortical limbic and brainstem structures. While these areas are clearly important for representing affect (e.g., valence and arousal), we propose that the default mode network (DMN) is additionally important for constructing discrete emotional experiences (of anger, fear, disgust, etc.). Findings from neuroimaging studies, invasive electrical stimulation studies, and lesion studies support this proposal. Importantly, our framework builds on a constructionist theory of emotion to explain how instances involving diverse physiological and behavioral patterns can be conceptualized as belonging to the same emotion category. We argue that this ability requires abstraction (from concrete features to broad mental categories), which the DMN is well positioned to support, and we make novel predictions from our proposed framework.
Keywords: abstraction; conceptualization; default mode network; emotion; memory; multivoxel pattern analysis.
Copyright © 2019 Elsevier Ltd. All rights reserved.