We offer an interdisciplinary framework for understanding the expanded capacity for emotional awareness (EA) in humans relative to other animals, synthesizing work within computational neuroscience, evolutionary psychology, and comparative anatomy. We argue that disproportionate cortical expansion during human evolution reflects additional hierarchical levels of computational processing, allowing representation of multimodal regularities over longer timescales - affording abstract concept learning, internal simulation of distal future outcomes, and expanded working memory capacity. This allows for the ability to simulate emotions, learn emotion concepts, and manipulate them in working memory when deciding how to act. We also draw on the construct of life history strategy within evolutionary psychology to argue that individual differences in EA within humans can be understood as the result of tuning particular computational parameters to the predictability of long timescale socioemotional regularities of the local environment. We conclude by discussing the implications and testable hypotheses offered by our proposed framework.