Humans reason and care about ethical issues, such as avoiding unnecessary harm. But what enables us to develop a moral capacity? This question dates back at least to ancient Greece and typically results in the traditional opposition between sentimentalism (the view that morality is mainly driven by socioaffective processes) and rationalism [the view that morality is mainly driven by (socio)cognitive processes or reason]. Here, we used multiple methods (eye-tracking and observations of expressive behaviors) to assess the role of both cognitive and socioaffective processes in infants' developing morality. We capitalized on the distinction between moral (e.g., harmful) and conventional (e.g., harmless) transgressions to investigate whether 18-mo-old infants understand actions as distinctively moral as opposed to merely disobedient or unexpected. All infants watched the same social scene, but based on prior verbal interactions, an actor's tearing apart of a picture (an act not intrinsically harmful) with a tool constituted either a conventional (wrong tool), a moral (producing harm), or no violation (correct tool). Infants' anticipatory looks differentiated between conventional and no violation conditions, suggesting that they processed the verbal interactions and built corresponding expectations. Importantly, infants showed a larger increase in pupil size (physiological arousal), and more expressions indicating empathic concern, in response to a moral than to a conventional violation. Thus, infants differentiated between harmful and harmless transgressions based solely on prior verbal interactions. Together, these convergent findings suggest that human infants' moral development is fostered by both sociocognitive (inferring harm) and socioaffective processes (empathic concern for others' welfare).
Keywords: human morality; infants; moral development; moral–conventional distinction; sympathy.