Two independent paradigms and methodologies (habituation and pupillometry) were used to record the surprise response of 7-month-olds 🤔when, in sequences of dyadic social interactions, we introduced a change in the appearance of the two interacting individuals (a boy and a girl)

Infants were tolerant to changes in body postures/movements, but not if those changes implied a change of roles. That is, they showed surprise when, say, the boy looked like the agent and the girl looked like the patient in trial n-1, but the opposite was true in trial n👇