Perceptions of moral character modulate the neural systems of reward during the trust game

Studies of reward learning have implicated the striatum as part of a neural circuit that guides and adjusts future behavior on the basis of reward feedback. Here we investigate whether prior social and moral information about potential trading partners affects this neural circuitry. Participants made risky choices about whether to trust hypothetical trading partners after having read vivid descriptions of life events indicating praiseworthy, neutral or suspect moral character. Despite equivalent reinforcement rates for all partners, participants were persistently more likely to make risky choices with the 'good' partner. As expected from previous studies, activation of the caudate nucleus differentiated between positive and negative feedback, but only for the 'neutral' partner. Notably, it did not do so for the 'good' partner and did so only weakly for the 'bad' partner, suggesting that prior social and moral perceptions can diminish reliance on feedback mechanisms in the neural circuitry of trial-and-error reward learning.