Abstract
Prior work has suggested that individuals with personality disorders (PD) have difficulties recognizing facial emotional expressions. However, most of this work has relied on small groups of clinical participants and targeted small sets of prototypical emotions, limiting the generalizability of the conclusions. Importantly, the extent to which PD-related impairments in self and interpersonal functioning underlie perception and memory of facial expression remain unclear. The present work aims to reveal associations of self and interpersonal functioning with pictorial representations of facial expressions using an image reconstruction methodology. We applied behavioural-based image reconstruction to a large set of facial expressions in 89 non-clinical participants. Similarity ratings of expressions for face stimuli and for faces recalled from memory were used to derive the appearance of perceptual- and memory-based representations. Further, a participant space was constructed and classification images were derived for its dimensions to investigate sources of individual variability in visual representations. Subsequently, participant scores on these dimensions were related to a test of facial recognition memory and measures of personality functioning (self and interpersonal) and pathological personality trait domains. Our findings revealed successful reconstructions of facial expressions from both perception and memory. Dimensions of individual variability in perceptual and memory representations of facial expressions accounted for differences in face recognition memory, as well as severity of several PD trait dysfunctions, including self-identity and empathy. Notably, expression reconstructions of individuals with poorer recognition abilities and greater personality impairments tended to be less accurate and expressive. In summary, the current work provides novel insights into individual differences in the representation of facial expressions, their relationships with PD-trait dysfunctions and face recognition, as well as their fine-grained visual content.