Imagine you observe a conflict between two co-workers, fellow team players, or cousins. Would you be puzzled if their boss, coach, or grandparent (respectively), who witnessed this conflict, did nothing about it? Most of us would answer 'yes' because (1) we identify such figures with a position of leadership, and (2) we associate leaders with a notion of responsibility.

Many social scientists believe that these expectations were shaped throughout human history. Our ancestors lived in small groups to survive, and leadership emerged to facilitate this group living. It was beneficial for individuals' survival to defer to those who took charge of the group (coordinating collective actions, resolving conflicts, and providing protection and care).

If this assumption is correct, and reasoning about leaders and their responsibilities toward their group is deeply rooted in our evolved cognition, then it need not be explicitly taught and could be evident early in life. In a series of studies, together with my collaborator, Dr. Renée Baillargeon, we chose ideal candidates to examine this possibility, infants. Inspired by research across several labs showing that infants are sensitive to social power asymmetries, we asked whether infants, similar to adults, specifically expect leaders to intervene and rectify transgressions they observe among their group members. Such behavior presents no direct benefits to the leader - if anything, an intervention can be costly. Thus, it could indicate that infants associate leaders with responsibility toward their group, and detect cases in which leaders shirk this responsibility, by not intervening.

To tackle such a loaded question, we used the common method of Violation of Expectation (VOE). This method takes advantage of infants' natural tendency to look longer at scenarios that are inconsistent rather than consistent with their expectations or knowledge. This method thus gives us insight into infants' reasoning. Specifically, in our studies, infants sat in front of a puppet stage and watched scenarios involving three bear puppets who served as the protagonist, the wrongdoer, and the victim. The protagonist brought in two toys for the other bears to share, but the wrongdoer seized both toys, leaving none for the victim. The protagonist then either took one toy away from the wrongdoer and gave it to the victim (intervention event) or approached each bear in turn without redistributing a toy (nonintervention event). Across conditions, the protagonist was either a leader (leader condition) or a non-leader equal in rank to the other bears (non-leader condition). Leadership was marked by either behavioral cues (the protagonist gave verbal instructions that exerted control over the other bears' behavior) or physical cues (the protagonist was larger than the other bears).

Infants in the leader condition looked significantly longer if shown the nonintervention as opposed to the intervention event. This behavior suggests that they had expected the leader to intervene and rectify the wrongdoer's fairness transgression, and were puzzled if the leader did not. In contrast, infants in the non-leader condition looked equally at the events. This behavior, in turn, suggests that they held no particular expectation for intervention from the non-leader. They could similarly make sense of a non-leader choosing the costly action of intervention, and choosing a safer action of nonintervention.

In a follow-up study, we found that infants' expectations concerning the leader's actions were nuanced and sensitive to the context that calls for its intervention. Infants were shown similar events, except for one critical change: when the leader protagonist brought in the two toys for the other bears to share, one bear displayed clear disinterest in the toys (uttered "No thanks", shook its head as in a no gesture, and looked down at its placemat), whereas, the other bear displayed an interest in obtaining the toys (uttered excitedly "Yay"), and subsequently seized both toys. From an adult perspective, this scenario could not be construed as a fairness transgression: One bear was interested in the toys, and the other was not. Seizing both toys here was, in fact, legitimate. Infants were then shown the same intervention and nonintervention events, as in the prior studies, with the leader protagonist either redistributing a toy or not, respectively. We found that infants looked longer at the intervention event than at the nonintervention event. This suggests that when no fairness transgression had occurred, they did not expect a leader to intervene.

Collectively, these studies show that by the second year of life, infants already ascribe unique responsibilities to leaders, which includes that of righting wrongs - well before they could be explicitly taught that leadership entails responsibilities.

Next read: Infants distinguish between leaders and bullies by Francesco Margoni

More from Psychology