Empathizing with basic emotions: Common and discrete neural substrates

Abstract

Empathizing is a quantitative trait involving understanding another's mental state (including their emotion) and responding to this with an appropriate emotion. A reliable, behaviorally validated self-report questionnaire measure of this is the Empathy Quotient (EQ), which is continuously distributed across the general population. The “discrete emotions” model posits that each “basic” emotion has a relatively independent evolutionary antecedent and social-communicative function and is subserved by a discrete neural system. In this study, we investigate if and how empathy influences the perception of basic emotions. Twenty-five volunteers (13 female, 12 male) selected across EQ space participated in a correlational design 3T fMRI study. The stimuli were presented in a box-car design, where 5 blocks (each containing 4 video clips of any one of happy, sad, angry, disgust or neutral expressions from different actors) and a low-level baseline were presented in pseudo-random order. Using an exploratory analysis, we found different brain regions correlated with EQ, depending on which emotion was being perceived. In particular, the ventral striatal response to happy faces correlated positively with EQ, while the ventral striatal response to sad faces was negatively correlated with EQ. The precuneus and lateral prefrontal cortical response to angry faces correlated positively with EQ. The response of the insula and the superior temporal gyrus cortex to disgust faces were negatively correlated with EQ. These results are discussed in the light of the postulated evolutionary function of each emotion. Using a hypothesis-driven conjunction analysis, we found that a region in the left dorsal inferior frontal gyrus/premotor cortex was positively correlated to the EQ across all four emotions. This region could therefore constitute a biomarker for trait empathy across emotions. We conclude that there are common regions underlying empathy across different emotions, and there are regions that show an emotion-specific correlation with empathy. This pattern of results is interpreted using a modification of Haxby et al.'s model of face perception.

Acknowledgments

SBC and ETB were funded by the MRC during the period of this work. The WBIC is supported by a MRC co-operative group grant. BC is supported by Trinity College, Cambridge.

Parts of this work were presented by BC at the Annual Meeting of the Society for Neurosciences, Washington, DC, USA, in November 2005, and submitted in part fulfillment of the degree of PhD by him. We are grateful to Carrie Allison and Sally Wheelwright for volunteer database management; John Suckling and Rhodri Cusack for support with fMRI analysis, Jac Billington for scanning support, Andrew Lawrence, Ilaria Minio-Paluello, Ofer Golan, Chris Ashwin and Chi-Hua Chen for valuable discussions, the Wolfson Brain Imaging Centre (WBIC), Cambridge for scanning facilities. We thank the two reviewers and the editor for their comments.