| The role of social influence in our lives cannot be overstated. Critical issues ranging from the outcomes of political campaigns to our stand on issues such as global warming, immigration and taxation depend on people trying to persuade the general public1. Social influence is also intertwined with our everyday life, as we try to persuade our children, influence our bosses and gain popularity among our friends. Research on social influence has been dominated by the motivation to understand the minds of the targets of influence—the “clients” (e.g., consumers and voters)—in order to exert even more influence on them2. Far less is known about the cognitive and neurobiological processes at play in the persuaders—the “advisers” (e.g., politicians and salesmen). Here we ask what happens in the advisor’s brain when engaged in the attempt to influence others. Bayarri and DeGroot3 proposed a normative solution for how advisers should modulate their strategy for offering their predictions in order to maximize their influence. They assumed that clients are affected by advisers’ accuracy and confidence4–7, i.e., they will be more likely to follow advisers that express confidence only when it is warranted and discredit highly confident but inaccurate advisers. In this case, in order to be selected by the client, advisers should modulate their advice-giving strategy depending on their current influence on the client. When the adviser is failing to influence the advisor, s/he should express higher confidence than permitted by objective evidence (positive advice confidence deviance). On the other hand, when the adviser is trying to maintain and protect already high influence, s/he should sit on the fence with cautious, nuanced advice that is lower than justified by the evidence (negative advice confidence deviance). We will call this a ‘competitive’ strategy for acquiring and maintaining influence. Social rank theory8,9 proposes an alternative account of advising behaviour by positing that people are not motivated just by the desire to influence others’ choices, but also by the fear of being excluded by the target of their influence, their client. This theory suggests that an adviser’s confidence should be proportional to his/her rank in a group. Lower rank individuals adopt submissive behaviours, expressing lower confidence and avoiding eye contact8. Accordingly, social rank theory suggests that, in contrast to the ‘competitive’ strategy described above, humans may adopt a ‘defensive’ strategy to manage influence by giving cautious advice when they are ignored by the client (i.e., when their influence is low) and exaggerating their confidence when their influence is high. In addition, social rank theory underscores the importance of an active process of social comparison by which an adviser evaluates her rank by tracking her performance relative to rival advisers10–14. In this view, relative performance or merit may also affect advice confidence leading people to display higher confidence when they think they perform better than their peers. |