This project investigates a question fundamental to social and economic interactions: Can individuals accurately assess the trustworthiness of unfamiliar others and, if so, through what mechanisms? Decisions to trust constitute a necessary element for the development of stable partnerships and societies. Yet, in the face of the many benefits that come with delayed exchanges, the decision to trust another is often a precarious one. The individual who first extends effort or provides resources is necessarily in the unenviable position of risking that the other will not reciprocate. Consequently, adaptive functioning relies not only on trusting others, but also on the ability to decide if one's partner is worthy of trust. Interestingly, however, previous efforts designed to uncover the signals of trustworthiness have been relatively unsuccessful. Building off initial research by members of the project team which demonstrates that trustworthiness of new partners can be discerned at greater than chance levels, the current project explores a novel route by which individuals may assesses the intentions of potential partners. Rather than looking for specific cues to trustworthiness in isolation, the process of assessing trust will be reconceptualized as a dynamic and iterative one. In short, the project will examine whether decisions to trust are not based simply on reading the cues of others, but rather on a dynamic "dance" whereby partners engage in nonconscious mimicry, and in so doing, are able to use perceptions of their own bodily states to infer their partner's feelings and, thereby, predict their motivations. To investigate this possibility, the research team will conduct experiments examining the interplay of emitted nonverbal signals and mimicry on decisions to trust within the context of behavioral economic games. Moreover, given that many facets of these phenomena often occur outside of human awareness and control, newly developed social robots will be used as interaction partners for humans in several experiments. These robots provide not only exacting control over relevant expressive parameters that is not possible in humans, but also the opportunity to test proposed models by allowing the robot to predict the trustworthiness of human partners. Taken as a whole, this project not only holds potential to increase understanding of the processes the underlie the initial emergence of trust-based relationships, but also to refine the physical and computational architecture that enable robots to infer motives and predict actions of humans and, thereby, increase their utility and value as interaction partners.
