Humans are, for the most part, moral. We behave in moral ways: We help and share with others, comfort those in distress, and cooperate with others to achieve far more than any one of us could achieve alone. We also have a sense of morality, which consists of thoughts and feelings about rights and duties, good and bad character traits, and right and wrong motives and behaviors (Krebs, 2008). Moreover, our moral apparatus contains moral emotions such as guilt and shame (Tangney, Stuewig, & Mashek, 2007). Together, these thoughts and feelings lead us to monitor, judge, and react to our own as well as to others’ motives and behaviors.

Humans demonstrate such morality in dyadic interactions, that is, when they are themselves engaged in and therefore directly affected by those interactions. If, for instance, someone steals from me, I might experience emotions such as anger or sadness, and I might retaliate against the person or try to punish him in some way. However, in dyadic interactions, the emotions and behaviors that look moral might actually be driven by selfinterest rather than by moral understanding as such. What is striking about human morality, however, is that we are also moral in third-party interactions, that is, when we are not directly engaged in or affected by the interactions but are simply observers of or aware of those interactions. Thus, if I observe someone stealing from someone else, I might again experience emotions such as anger or moral indignation, and I might retaliate against the person or try to have the person punished. Of importance, morality in these types of interactions cannot simply be driven by self-interest but must be driven at least in part by an understanding of moral norms in an agent-neutral manner, that is, as they apply to all people rather than to me alone.

Such third-party interactions are considered the litmus test of moral understanding because they tap into agent-neutral (rather than self-interested) applications of moral norms. Philosophers such as Hume, Kant, and Rawls, for instance, have argued that the essence of true morality is agentneutral morality, which is achieved by moving away from an egocentric perspective and assuming an impartial point of view. Thirdparty intervention is also considered essential in evolutionary analyses of the origins of morality because it is argued that intervention in dyadic interactions alone would likely not sustain the kind of largescale cooperation seen among humans (Boyd & Richerson, 1992; Fehr & Fischbacher, 2003). Indeed, third-party intervention is thought to be unique to humans (e.g., Krebs, 2008; Tomasello, 2009). It is thus critical to investigate the ontogeny of morality in third-party contexts.

The papers in this dissertation thus explored the ontogenetic emergence of some critical aspects of agent-neutral morality. Specifically, in four studies, I examined young children’s understanding of and their responses to third-party moral transgressions and the victims and transgressors therein.

Study 1 explored 18- and 25-month old children’s sympathetic and prosocial responses to a victim (Vaish, Carpenter, & Tomasello, 2009). In most research on the early ontogeny of sympathy, young children are presented with an overtly distressed person and their sympathetic responses are observed. This work leaves unclear whether, when the victim’s distress is not perceptible, young children can nevertheless sympathize with the victim. Children in Study 1 saw an adult either harming another adult by destroying or taking away her possessions (harm condition) or else doing something similar that did not harm her (neutral condition). The “victim” expressed no emotions in either condition. Nevertheless, in the harm as compared with the neutral condition, children showed more concern and subsequent prosocial behavior towards the victim. Children’s concerned looks during the harmful event were also positively correlated with their subsequent prosocial behavior towards the victim. Thus, by 18 months of age, children can sympathize with the victims of third-party moral transgressions even in the absence of overt emotional cues from the victims, possibly by some form of affective perspective- taking, and they subsequently behave prosocially towards the victims of such transgressions. Studies 2 and 3 assessed various forms of third-party intervention. The two parts of Study 2 assessed third-party intervention in the form of selective prosocial behavior towards actors in third-party interactions (Vaish, Carpenter, & Tomasello, 2010). In Study 2a, 3- year-old children watched one adult (the actor) harming or helping another adult. Children later helped the harmful actor less often than a third (previously neutral) adult, but helped the helpful and neutral adults equally often. In Study 2b, 3-year-old children helped an actor who intended but failed to harm another adult less often than a neutral adult, but helped an accidentally harmful and a neutral adult equally often. Young children thus selectively avoid helping those who cause – or even intend to cause – others harm.

Study 3 examined whether children actively enforce agentneutral moral norms by intervening in and attempting to prevent third-party moral transgressions (Vaish, Missana, & Tomasello,  2011). Three-year-old children and two puppets each created a picture or clay sculpture, after which one puppet left the room. In the harm condition, the remaining (actor) puppet then destroyed the absent (recipient) puppet’s picture or sculpture. In a control condition, the actor acted similarly but did not harm the recipient. Children protested during the actor’s actions (often using normative or moral language), and, upon the recipient’s return, tattled to the recipient about the actor and behaved prosocially towards the recipient more in the harm than in the control condition. Together, Studies 2 and 3 show that young children actively intervene in moral transgressions (by protesting), demonstrate at least some forms of punishment or shunning of moral transgressors (by withdrawing help from them and tattling on them), and act prosocially towards victims even when those victims were not present and thus did not provide any emotional cues during the transgressions (extending the findings of Study 1).

Study 4 examined the flexibility of children’s moral understanding (Vaish, Carpenter, & Tomasello, 2011). I asked whether 4- and 5-year-old children, who have previously been shown to make quite sophisticated moral judgments about transgressors, have the flexibility to modify those judgments based upon the transgressor’s subsequent remorse. After children watched videos of transgressors either displaying or not displaying guilt, 5-year-olds appropriately inferred that the victim would be more upset with the transgressor who had not displayed guilt and would prefer the guilt-displaying transgressor. The 5-year-olds also said that they would prefer to interact with the guilt-displaying transgressor, judged the transgressor who had not displayed guilt to be meaner, and, in a distribution of resources task, gave more resources to the guilt-displaying transgressor. The 4-year-olds did not draw any of these inferences and distributed the resources equally to the two transgressors. Thus, between 4 and 5 years of age, children’s moral judgments and behaviors become flexible enough to vary based upon whether or not the transgressor displayed remorse. Furthermore, this study charted for the first time the development of an understanding of the appeasement functions that displaying guilt (and other social emotions) serves in restoring cooperative relationships and thus in maintaining cooperation in groups (e.g., Keltner & Anderson, 2000). Together, the studies in this dissertation point in several ways to a sophisticated and flexible morality early in human ontogeny, as evident in children’s moral emotions, behavior, and judgments. First, they demonstrate the flexible nature of empathy-related processes even in young children. As noted by Hoffman (2000), empathic arousal is a reliable prosocial motivator because it is multidetermined: It can be elicited by observing clear and overt distress signals as well as by imagining the other’s distress. This multideterminism adds scope to one’s empathic capability. Indeed, adults can sympathize with victims not only when they have direct access to the victims’ distress cues but even when they do not (e.g., Batson et al., 1991; Decety & Jackson, 2004). The findings of Study 1 show that this multi-determined nature of empathy-related responses is already functional in early development. This suggests that even in early ontogeny, empathy-related responses are reliable proximate mechanisms underlying moral behavior.

Relatedly, the studies in this dissertation show that even early in development, morality is not rigid and driven by simple rules such as “Always help people who need help” or “People who cause harm are always bad.” Rather, early morality seems to be sophisticated, and if it is driven by rules, then it at least seems to be driven by complex and multifaceted rules. The evidence for this was manifold across studies. First, it emerged that children’s prosocial behavior is selective. Thus, at 18 months, 2 years, and 3 years of age, children are more prosocial towards a victim than a non-victim (Studies 1 and 3). At 3 years of age, children reduce their helping towards transgressors (Study 2), and 5-year-old children selectively help a remorseful transgressor over an unremorseful one (Study 4).

Early moral flexibility was also revealed in terms of the factors that young children took into account in their moral evaluations and behaviors. Study 2a showed that 3- year-olds understand basic moral transgressions and they selectively reduce their helping behavior towards transgressors. Study 2b further showed that 3-year-olds take into account a transgressor’s intentions, only reducing their prosocial behavior towards a transgressor who intended but was unable to cause harm, not towards a transgressor who unintentionally caused harm. Moreover, Study 4 revealed that children also take a transgressor’s subsequent reaction (remorse or no remorse) into account in their moral evaluations and behaviors. All in all, then, the studies in this dissertation provided strong evidence for flexible moral understanding in early ontogeny.

A final major finding of this dissertation was that young children are deeply concerned with and involved in third-party moral interactions. In all four studies, children were witnesses to (rather than active participants in) moral transgressions. Yet they not only paid close attention to the interactions but they also became actively involved in those interactions, affectively, behaviorally, and through their evaluations. The most striking finding in this regard was that even young children demonstrated some forms of third-party punishment. For example, when given a choice between instrumentally helping a perpetrator and a neutral person, 3-year-old children generally chose not to help the perpetrator (Studies 2a and 2b). Children thus engaged in a common form of third-party punishment, namely, withdrawing cooperation from perpetrators (Boyd & Richerson, 2005). Two other forms of thirdparty intervention emerged in Study 3: (i) When a victim was absent during a moral transgression, 3-year-old children actively intervened (by verbally protesting) against the transgression, presumably to prevent it from occurring, and (ii) upon the victim’s return, they tattled to the victim about the perpetrator, perhaps as a way to ensure that the perpetrator would be punished. As mentioned previously, involvement and intervention in third-party transgressions are thought to provide strong evidence for moral understanding because they demonstrate agentneutral applications of moral norms. Third-party intervention is also considered essential in evolutionary analyses of the origins of morality. The findings of this dissertation demonstrate that an essential and perhaps unique aspect of human morality, third-party intervention, is already evident in early ontogeny.

Together, the studies in this dissertation contribute to our understanding of the emergence of morality: They point to deep ontogenetic roots of human morality, provide evidence for a flexible and sophisticated morality in young children, and demonstrate the early ontogenetic emergence of a potentially human-unique aspect of morality, namely, agent-neutral morality.

References

Batson, C. D., Batson, J. G., Slingsby, J. K., Harrell, K. L., Peekna,

H. M., & Todd, R. M. (1991). Empathic joy and the empathyaltruism hypothesis. Journal of Personality and Social Psychology, 61(3), 413-426.

Boyd, R., & Richerson, P. J. (1992). Punishment allows the evolution of cooperation (or anything else) in sizable groups. Ethology and Sociobiology, 13, 171-195.

Boyd, R., & Richerson, P. J. (2005). Solving the puzzle of human cooperation. In S. Levinson (Ed.), Evolution and culture (pp. 105-132). Cambridge, MA: MIT Press.

Decety, J., & Jackson, P. L. (2004). The functional architecture of human empathy. Behavioral and Cognitive Neuroscience Reviews, 3(2), 71-100.

Fehr, E., & Fischbacher, U. (2003). The nature of human altruism. Nature, 425, 785-791.

Hoffman, M. L. (2000). Empathy and moral development: Implications for caring and justice. Cambridge, UK: Cambridge University Press.

Keltner, D., & Anderson, C. (2000). Saving face for Darwin: The functions and uses of embarrassment. Current Directions in Psychological Science, 9(6), 187- 192.

Krebs, D. L. (2008). Morality: An evolutionary account. Perspectives on Psychological Science, 3 (3), 149-172.

Smetana, J. G. (2006). Socialcognitive domain theory: Consistencies and variations in children's moral and social judgments.

In M. Killen & J. Smetana (Eds.), Handbook of Moral Development (pp. 119-153). Mahwah, NJ: Lawrence Erlbaum Associates.

Tangney, J. P., Stuewig, J., & Mashek, D. J. (2007). Moral emotions and moral behavior. Annual Review of Psychology, 58, 345-372.

Tomasello, M. (2009). Why we cooperate. Cambridge, MA: MIT Press.

Vaish, A., Carpenter, M., & Tomasello, M. (2009). Sympathy through affective perspectivetaking and its relation to prosocial behavior in toddlers. Developmental Psychology, 45(2), 534- 543.

Vaish, A., Carpenter, M., & Tomasello, M. (2010). Young children selectively avoid helping people with harmful intentions. Child Development, 81(6), 1661- 1669.

Vaish, A., Carpenter, M., & Tomasello, M. (2011). Young children's responses to guilt displays. Developmental Psychology, 47(5), 1248-1262.

Vaish, A., Missana, M., & Tomasello, M. (2011). Three-yearold children intervene in thirdparty moral transgressions. British Journal of Developmental Psychology, 29(1), 124-130.

Warneken, F., & Tomasello, M. (2009). The roots of human altruism. British Journal of Psychology, 100(3), 455-471.