The Origins of Morality: Evidence that some aspects of morality are innate


The concept of innate knowledge has incited much dispute among philosophers and cognitive scientists alike. Recent methodological advances have allowed developmental researchers to empirically investigate what innate knowledge humans might actually possess, and how such core knowledge affects infants’ construction of information as they develop and interact with the world. This article focuses on the domain of morality – though morality is not easily defined, it can be understood as a system of values and principles concerning the distinction between right and wrong behavior. Morality, then, requires the ability to evaluate the intentions, goals, mental states, and behavior of others; it involves a sense of fairness and a sense of justice when it comes to the distribution of resources, rewards, and punishments, and sensitivity to the well-being of others (exhibited by empathy), which prompts a variety of prosocial and altruistic behaviors (sharing, helping, cooperating, and comforting others in distress). This article reviews research with infants and toddlers that has demonstrated that from very early in life, humans exhibit a wide range of moral behaviors and judgments: infants evaluate others’ behavior, they are able to distinguish between helpful agents and harmful ones, and they prefer helpful agents over harmful ones; infants collaborate with others to achieve goals, they empathize with others in distress, and have rudimentary senses of fairness and justice. Given that these abilities are present so early in life and can also be observed in other social animals, it is unlikely that these behaviors and judgments emerge solely as the result of socialization and learning. It seems more likely that infants have core knowledge of morality, and that this moral sense evolved to facilitate the cooperative behaviors necessary for successful group living.


Where does our ability to determine right from wrong come from? Why is it that we feel that certain behaviors, acts, or individuals are evil, and that others are good?

Explorations into the origins of morality have long been left to the realms of theology and philosophy, yet in recent decades, psychologists have sought to answer this question empirically by studying moral behaviors and reasoning in humans, as well as by examining similar behaviors in other social animals. Humans are social animals; in order to survive, we must get along and help one another. Our survival depends on our ability to cooperate. Cooperation requires individuals to either equate their self-interests with those of others (based on notions of empathy, fairness, equality, justice, or reciprocity) or to suppress those interests (to help and to share). From an evolutionary standpoint, morality can be understood as a system of motivations and abilities for cooperation. Our sense of morality prompts us to cooperate with others, and regulates our social interactions in a way that helps groups to live together and thrive.

Previous research on the development of morality has speculated that a sense of morality develops over time as the result of parental instruction and socialization, through which children learn what constitutes socially acceptable behavior and gradually build moral codes based on their own interactions and observations (see Spinrad et al., 2006; Turiel, 2006; for reviews). Yet, this explanation fails to consider the role intuitions play in the development of morality, and mounting evidence that certain moral senses exist very early in life renders explanations of moral development that rely solely on learning insufficient. It is far more plausible that moral development builds upon early-emerging social or moral senses about how individuals should treat each other (Dupoux & Jacob, 2007; Greene, 2005; Haidt, 2001, 2008).

In this paper, I will review the literature on moral behavior and reasoning in infancy, as well as some literature on moral behavior in other social animals. I will review studies which demonstrate that infants are capable of social evaluation (e.g. preferring helpers to hinderers; Hamlin, Wynn, Bloom, 2007; Kuhlmeier et. al, 2003; Premack & Premack, 1997) and studies which show that infants have a complex sense of fairness in regards to the distribution of resources (Geraci & Surian, 2011; Sloane, Baillargeon, & Premack, 2012). I will also review literature that examines displays of empathy in young infants (Zann-Waxler et al., 1992), which often prompts their prosocial behaviors. I will argue that the literature suggests that infants do indeed appear to possess an intuitive moral sense, and provides evidence for the theory that humans possess core knowledge of morality. I will further argue that the existence of similar moral evaluations and behaviors in our primate relatives (e.g. the capacity for social evaluation, a sense of fairness, and prosocial behavior) supports the notion that some aspects of morality are innate and can be explained as having evolved to facilitate cooperation within social groups (Melis et al., 2006; Warneken & Tomasello, 2006, 2009; Brosnan, 2003).

Social evaluation

If we are to understand our moral sense as having evolved to promote the cooperative behaviors necessary for successful group living, it would make sense that our moral system would require a sense of fairness, or a sensitivity to equality in regard to resource distribution. The ability to determine fairness when distributing resources also requires the ability to distinguish those who contribute from those who do not, or an evaluative capacity to identify those who are helpful and cooperative from those who are unhelpful and uncooperative, as well as a desire to punish or dislike the latter.

Research has shown that infants engage in cooperative interactions during their first year of life (Bates et al., 1979; Ross & Lollis, 1987). For example, infants coordinate their actions to those of a social partner in cooperative routines (e.g. peak-a-boo) before their first birthday. Evidence has also shown that during the following years (roughly 13 to 30 months), infants become more skilled collaborative partners and become capable of engaging in complementary actions with a social partner in a variety of cooperative activities over the course of the following years. (Brownell & Carriger, 1990; Brownell et al., 2006; Eckerman et al., 1989; Eckerman & Didow, 1989; Warneken & Tomasello, 2007; Warneken et al., 2006). However, exactly how much infants understood about collaborative goals remained unclear. A recent study by Henderson and Woodward (2011) sought to investigate infants’ understanding of the collaborative-goal structure of collaborative actions, and asked whether 14-month-old infants are able to go beyond analyzing the goals of single individuals when they observe two individuals produce complementary actions in order to achieve a collaborative goal.

The researchers used a visual habituation paradigm to assess whether infants understood that collaboration requires that the actions of cooperative partners are both (1) complementary and critical to goal attainment and (2) driven by a shared intention to attain a common goal. During habituation trials, infants saw an event in which one agent (the box-opener) used both hands to retrieve the box, then examined the box and opened the lid. Next, the other agent (the duck-getter) used both hands to retrieve the duck, then played with the duck. Both agents smiled at each other and the box-opener closed the box. The agents smiled at each other again and the trial ended with both agents looking down at the object they had acted on. This sequence of events was the same in both the collaboration and the no-collaboration condition, except that in the no-collaboration condition, the duck was beside, not inside, the box. During test trials, the box and the duck were placed equidistant from one another on a stage in front of the box-opener. The infants watched a series of test events in which the box-opener alternated grasping the box or the duck.

The question of interest was whether infants inferred that the goal of box-opener was to open the box, or if infants inferred that the box-opener’s goal was to retrieve the duck. If infants thought that the box-opener’s goal was the box, they should look longer when the agent unexpectedly reaches for the duck. Similarly, if infants thought that the box-opener’s goal was the duck, they should look longer when the agent reaches for the box. The analysis of infants’ looking times from the collaboration condition revealed longer overall looking times during the box test trials (when the box-opener reached for the box). Thus, infants who watched the collaborative event inferred that the box-opener’s goal during habituation trials was retrieving the duck. Infants who watched the agents act independently in the no-collaboration condition did not show a reliable response to the box-opener’s goal. Henderson and Woodward’s results suggest that 14-month-old infants who viewed the collaborative event interpreted the event in terms of a collaborative goal. The infants looked longer in the test trials when the box-opener reached for the box than when the box-opener reached for the duck, despite the fact that the agent had never touched the duck during the habituation trials. Additionally, infants who viewed the non-collaborative event in which agents acted independently on either the box or the duck did not differentiate between the test events. Their evidence suggests that by 14 months, infants understand that the actions of agents engaged in collaboration is complementary and essential to attaining their goal, and that their actions are undertaken in order to achieve a shared goal. Thus, it is clear that from very early in life, infants are capable not only of engaging in collaboration with others, but also of understanding how the actions of other agents may be coordinated in order to attain a common goal.

While Henderson and Woodward’s work demonstrates that infants understand cooperative behaviors, other research has sought to understand how infants evaluate others based on their behaviors. Previous research has shown that when an infant sees an object moving spontaneously and displaying goal-directed action, the infant will interpret the object as intentional and attribute psychological properties to the object (Premack & Premack, 1990; Biro et al., 2007). A study by Premack and Premack (1997) investigated whether infants attribute value to interactions between intentional objects. Infants (12 months of age, on average) were shown computer-generated animations of spontaneously moving balls. There were two positive interactions between the balls – one ball either “caressed” another, or one ball “helped” another by moving it towards its goal. There were also two negative interactions, in which one ball either “hit” another, or “hindered” it, by preventing another ball from reaching its goal. The researchers used a habituation/dishabituation paradigm and their measures of infants’ looking times revealed that only the infants who had been habituated to a positive condition dishabituated when they viewed a negative condition, and only infants who had been habituated to a negative condition dishabituated when they were transferred to a positive condition. This shows that infants equated a gentle action with a helping action, and also equated a harmful action with a hindering action, despite the fact that these acts were not physically similar. Thus, by 12 months of age, infants attribute value to the interaction of intentional objects: they assign positive value to gentle actions and to goal-helping behavior, and negative value to harmful actions and to goal-hindering behavior.

Premack and Premack’s work demonstrates that 12-month-old infants evaluate intentional objects based on their actions toward other objects, and a study by Hamlin, Wynn, and Bloom (2007) went further by investigating whether 6 and 10-month-old infants evaluate agents based on their actions towards others, and whether the infants form preferences for the agents based on their observations. In their first experiment, infants saw a character (‘climber,’ a block with googly eyes) sitting at the bottom of a hill. During habituation, infants saw the climber attempt to climb the hill, and on the third try, the climber was either helped up the hill by a helper pushing it up from behind, or the climber was pushed down the hill by a hinderer. Next, infants’ preferences were evaluated by giving them the choice to choose between the two blocks (helper and hinderer) and recording which one infants reached for. The choice measure revealed that infants preferred to reach for the helper. This suggests that infants generate preferences for agents based upon their actions towards others.

In experiment 2, Hamlin, Wynn and Bloom sought to assure that infants’ behavior in experiment 1 was really due to social evaluation and not to some general perceptual preferences for upward or downward movement. They showed infants identical events as in experiment 1, except that the blocks did not have eyes. Their results revealed that neither the 10-month-olds or the 6-month-olds robustly preferred the object that pushed up to the object that pushed down. The fact that infants overwhelmingly preferred to reach for the block who pushed up when this action occurred in a social context (when the block was an agent) indicates that infants’ generated preferences based upon their social evaluations of the events, and not on perceptual differences between the events.

The infants’ choice patterns indicate three possibilities: infants may negatively evaluate an agent they observe hindering another (thus find the hinderer aversive); they may positively evaluate an agent they observe helping another (thus find the helper appealing); or they may use both negative and positive evaluation processes. In order to better understand what evaluations infants were basing their preferences on, Hamlin, Wynn, and Bloom conducted a third experiment which included a neutral character. In experiment 3, infants were habituated to either a helper or a hinderer acting on a climber, just as in experiment 1, but they also saw a neutral block moving upward or downward on the hill. The movements of the neutral block were identical to that of either the helper or the hindered except that the movements of the neutral block did not interact with the climber whatsoever. The researchers then tested infants’ preferential reaches, and found that both the 6-month-old infants and the 10-month-old infants responded differently to the neutral block depending upon whether they had to choose between reaching for the neutral block or the helper or between the neutral block or the hinderer. They found that infants reached for the helper over the neutral agent, and that they reached for the neutral agent over the hinderer. This shows that infants not only avoid harmful agents, they also prefer helping agents. Their findings suggest that preverbal infants prefer helping agents to hindering ones, prefer helping agents to neutral agents, and prefer neutral agents to hindering agents. Two evaluative mechanisms are at work here – a preference for prosocial behavior and an aversion to antisocial behavior. The nuance of very young infants’ evaluative capacities lend evidence to the notion that some aspects of morality are innate, as it is unlikely that infants’ consistent preferences for helpers over hinderers is the result of socialization or learning this early in life.

Another piece of adult morality is a sense of retribution or justice, which can be understood as the desire to execute or support the punishment of those who are uncooperative or act as hinderers. A sense of justice also includes the desire to reward those who do cooperate and are helpful. It would make sense that a system of morality would require not only the ability to evaluate the behavior of others, but also a desire to respond accurately to those behaviors of other group members, if we are to understand our moral sense as having evolved to facilitate cooperation. Hamlin, Wynn, Bloom, and Mahajan (2011) asked whether 21-month-old infants exhibited any rudimentary sense of retribution. In this experiment, toddlers were shown a scenario in which a puppet attempted to achieve a goal (e.g. opening a box) and another puppet either helped the first puppet achieve the goal, or prevented it (e.g. by slamming the box closed). Then toddlers were given the opportunity to either reward (by giving a treat to) or punish (by taking a treat from) the puppets. Their results showed that the children preferred to give to the helpful puppet and preferred to take from the unhelpful puppet.

Hamlin et al. (2011) then asked whether even younger infants were sensitive to the fairness of retribution behavior. In their first experiment, the researchers examined whether infants positively evaluate an agent who acts negatively toward another agent who had just acted antisocially. To do this, they showed 5 and 8-month old infants a scenario in which one puppet repeatedly attempted to open a box. Then they saw a trial in which another puppet (prosocial puppet) helped the first puppet open the box, and another trial in which infants saw an antisocial puppet hinder the first puppet by slamming the box closed. After observing these interactions, infants were placed in either the prosocial target condition or the antisocial target condition. In the prosocial target condition, infants watched the prosocial puppet from the previous task playing with a ball, which the puppet then dropped. In separate trials, the prosocial puppet’s ball was either given back by another (the giver) or taken away by another (the taker). In the antisocial target condition, infants watched the same events (puppet drops ball and ball is either returned or taken away) except that the antisocial puppet was the puppet playing with the ball. After observing these events, the infants were then presented with the giver and the taker and encouraged to choose one of them. Their results showed that when the target of the actions was the puppet infants had just observed acting prosocially, infants preferred to reach for the giver. Infants’ preferences were more nuanced however, when it came to reaching for either the giver or the taker when the target of the actions was the antisocial puppet. Five-month-olds preferred to reach for the giver, while 8-month-olds preferred to reach for the taker. These results demonstrate the although younger infants uniformly prefer agents who behave positively or prosocially, by 8 months of age, infants selectively prefer agents who behaved positively toward prosocial agents and agents who behaved negatively toward antisocial agents. This adds to previous evidence that infants’ are capable of making complex social evaluations from very early in life.

Social evaluation in non-human animals

In humans, a capacity for social evaluation and a preference for cooperative individuals emerge very early in life. The early emergence of these abilities lends support to the theory that such abilities and preferences are innate, not learned. The ability to evaluate others based on their actions and a preference for more cooperative individuals can also be observed in other social animals, and this lends further support to the notion that these moral senses are evolutionarily ancient and likely evolved to promote the cooperative behaviors necessary for successful group living.

Observations of chimpanzees in the wild show that chimps seem to coordinate their positions in trees during hunts in order to surround their prey (Boesch, 2000; Watts & Mitani, 2002). However, the cognitive skills behind such cooperative activities remain unclear. A study by Melis et al. (2006) investigated whether chimpanzees know when collaboration is necessary and whether they know to choose the more effective of two potential collaborative partners, based on their previous experience with each partner. In their experiment, eight semi-free-ranging chimpanzees were given the opportunity to recruit a collaborative partner when they either needed help to retrieve the food or did not need help to retrieve the food. The chimps were introduced a feeding platform (a platform with food resting on top) in separate sessions, and the platform was placed out of reach of the subjects. A rope was attached to the feeding platform and both ends of the rope extended into the testing room. If the chimps pulled on both ends of the rope simultaneously, the feeding platform would move toward the chimps and become reachable. If only one end of the rope was pulled, the rope became unthreaded and thus the food remained unreachable. In the collaboration condition, the subject chimp and the partner chimp watched from separate rooms as the rope attached to the feeding platform was positioned so that the two ends of the rope were too far apart for one individual to grasp simultaneously. The subject was then released into the testing room, while the partner remained locked in an adjacent room that only the subject could open. In the solo condition, the ends of the rope were close enough that one chimp could pull both ends at once without help. The researchers found that subjects unlocked the door to recruit their partner significantly more often in the collaboration condition. This demonstrates that chimps are capable of understanding when they need to cooperate with another individual to complete a task.

Given that chimpanzees tend to recruit a collaborator only when needed, Melis et al. conducted a second experiment to test whether chimps can also learn to recruit the more effective of two partners based on their previous interactions. Six chimpanzees that had participated in the first experiment were placed in the testing room. Two potential collaborators were placed in the two rooms adjacent to the testing room. The subject in the testing room could easily unlock the door to either of the two adjacent rooms. The two potential collaborators had demonstrated very different levels of success in the previous experiment when working with the subject to pull the feeding tray within reach. Thus, one potential collaborator was a more effective partner and the other was a less effective partner. The testing procedure was identical to that of the collaborative condition from the first experiment, except that there were now two potential partners locked in two separate rooms. The results of this experiment revealed that the subjects overwhelmingly preferred to recruit the more effective partner. This demonstrates that that chimps are capable of evaluating others based on their previous actions, and also that chimps act in accordance with these evaluations, as they tended to choose the more effective partner.

Fairness when distributing resources and rewards

From an evolutionary perspective, a moral sense that evolved to promote cooperative behavior would understandably require sensitivity to the equitable distribution of resources. The studies above have made it clear that infants are capable of evaluating the actions of others from very early in development. Here I will discuss more studies which demonstrate that infants have a complex sense of fairness from very early in life: by the second year of life, infants expect agents to treat others fairly (Geraci & Surian, 2011) provided that all agents have contributed equally to the cooperative task (Sloane, Baillargeon, & Premack, 2012).

In order to assess whether infants can take into account the outcome of distributive actions when evaluating and reasoning about agents’ actions, Geraci and Surian (2011) examined 12 to 18 month-old infants’ responses to events involving equal and unequal resource distribution. Infants saw animated events in which one agent distributed resources equally between two recipients, and another agent distributed resources unequally between two recipients, while a separate bystander observed all of the distributive actions. Next, infants saw the bystander begin to approach one of the distributors, and infants’ anticipatory looks to either the fair or the unfair distributor were recorded. Geraci and Surian found that infants expected the bystander to approach the fair distributor. Then, infants were presented with two pictures, one of the fair distributor and one of the unfair distributor. Infants preferentially reached for the fair distributor. The results of the experiment reveal that by 12 to 18 months of age, infants possess some rudimentary notions of fairness: infants evaluated agents based on their distributive actions, preferred the fair distributor to the unfair distributor, and reasoned that other agents (within the animation) should likewise prefer the fair distributor.

Infants prefer agents who distribute resources fairly, and they also have expectations about how resources should be distributed. In order to examine infants’ expectations about how agents should distribute resources and rewards to other agents, Sloane, Baillargeon and Premack (2012) conducted an experiment in which 19-month-old infants watched an experimenter distribute resources to two puppets. In the experiment, the infants saw an experimenter divide two desirable items equally or unequally between two puppets, who danced and clapped and responded excitedly when the experimenter showed them the items (two toy cars, two edible cookies, or two toy ducks). The researchers measured looking times to see if infants would detect a violation when the experimenter distributed resources unequally. The results of Experiment 1 showed that infants looked longer when the experimenter distributed resources unequally, which suggests that 19-month-old infants expect others to distribute resources equally between two similar individuals.

What’s more, the results of Experiment 1 show that it is unlikely that the infants’ expectations reflects some low-level cognitive factors, given that infants did not display this expectation when the puppets were inanimate. Everything in the inanimate-control condition was identical to the experimental condition except that the puppets did not move or talk. Measures of infants’ looking times revealed that infants looked equally at both the unfair distribution and the fair distribution events when the puppets were inanimate. Further, infants did not simply expect that similar individuals should have a similar number of items. In the cover-control condition, two opaque boxes were in front of each puppet, and the experimenter removed the box to reveal either one object in front of each agent, or two objects in front of one agent and none in front of the other. This condition removed the role of the distributor, and measures of infants’ looking times again revealed that infants looked equally at both events. This suggests that infants did not simply expect that similar individuals should have a similar number of objects; their expectation has to do with the distribution. Infants in the inanimate condition and in the cover condition looked equally at the two events, which provides further evidence for the claim that 19-month-olds expect a distributor to divide resources equally between two individuals.

Sloane, Baillargeon and Premack conducted a second experiment in order to determine whether infants held more nuanced expectations about the way rewards should be distributed, namely, if infants expected that agents should be rewarded in accordance with the amount of effort they put in to a task. In Experiment 2, 21-month-old infants watched an experimenter ask two participants (human agents, no more puppets) to put away toys. In the explicit condition, the experimenter told the participants that they would receive a reward if they put away the toys, while in the implicit condition, the experimenter did not mention rewards, only asked the participants to put away toys. This was to see whether infants had expectations regarding the distribution of rewards even without an explicit contract. In both conditions, infants watched either one participant working (the worker) to put the toys away while the other (the slacker) played, or both participants working. The experimenter left after telling the participants to put away the toys, and returned after the toys were put away in both the one-works event and the both-work event. The experimenter then looked at the transparent box in front of each participant in order to determine how much work they had done (in the one-works condition, all of the toys were in the transparent box in front of the worker, while in the both-work, the toys had been placed equally in both boxes). Next, the experimenter gave a reward (a sticker) to both participants. The results revealed that infants looked longer in the one-works event in both the implicit and explicit conditions. This suggests that 21-month-old infants expect a distributor to reward individuals in accordance with their efforts. Given that infants still looked longer when the worker and the slacker were rewarded equally even in the absence of any promise of reward for completing the work, it is clear that no prior explicit contract is necessary in order for infants to have expectations regarding the distribution of rewards.

The control condition of the experiment ruled out the possibility that infants looked longer at the one-works event not because they were confused or surprised that the experimenter rewarded the worker and the slacker equally, but because they were responding to peripheral aspects of the events (e.g., they preferred to see all of the toys in one box, or they were confused that the slacker did not help). All of the test events in the control condition were identical, except that the boxes that the participants placed the toys in were opaque, and therefore the experimenter was unable to determine whether both participants had worked or only one had. In this case, infants looked about equally at the one-works and both-work events. This demonstrates that infants held expectations about the actions of the experimenter only when the experimenter could determine who had worked and who had not, as infants did not look longer when the experimenter rewarded the worker and the slacker equally when the experimenter could not see the contents of the boxes. Taken together, the results of these experiments demonstrate that by the second year of life, infants possess context-sensitive expectations about the fair distribution of resources and rewards.

Fairness when distributing resources and rewards in non-human animals

In humans, a preference for equally distributing collaboratively earned resources and for fairly distributing punishment emerges very early in development. The presence of these senses so early in life lends support to the notion that such abilities are innate. The presence of similar preferences for fairness in other social animals provides further evidence for this claim, and also lends support to the notion that that such moral senses are likely the result of evolutionary development favoring fair and cooperative interactions among individuals that facilitate successful group living.

Negative reactions to unequal reward distribution have also been reported in others species, including brown capuchin monkeys (Brosnan & de Waal, 2003), chimpanzees (Brosnan, Schiff & de Waal, 2005) and dogs (Range, Horn, Viranyi & Huber, 2009). Brosnan & de Waal’s study (2003) sought to examine how nonhuman primates (brown capuchin monkeys) respond to unequal reward distribution in exchanges with human experimenters. In the experiment, two monkeys in cages next to each other had to hand the experimenter a rock in order to receive a reward. One monkey was then given a cucumber as a reward while the other monkey, who performed the same exact task, received a grape. Monkeys who received the less desirable reward (the cucumber) were more likely to react negatively, such as by throwing the reward back at the experimenter, beating their arms against the cage in anger, or by refusing to continue participating in the exchanges. Additionally, the experimenters found that the monkeys were even more likely to reject the reward as the situation became increasingly unfair. In an effort-control condition, monkeys saw their partner receive a grape without exchanging a token for the grape. In this case, monkeys were more likely to refuse the cucumber reward or react negatively than they were when they had simply received the less desirable of the two rewards. These findings suggest that monkeys have a rudimentary sense of fairness, as monkeys display behavior that indicates they are averse to inequity.

A study by Brosnan et al. (2010) sought to better understand nonhuman primates responses to inequity. The researchers used the same procedure to test chimpanzees that Brosnan and de Waal (2003) used to test capuchin monkeys. The chimps were trained to exchange an inedible token for a food reward, and then the chimps were tested in pairs as they sat in cages next to each other. When the chimp handed the token to the experimenter, the experimenter held it up in front of the chimpanzee, then held up the reward (a grape was the high-value reward, a carrot was the low-value reward) so that both chimpanzees could see it and gave the reward to the chimpanzee who had just completed the exchange. Their results provided further evidence that nonhuman primates respond negatively to inequity, as the chimpanzees in the study refused to complete the exchange or refused to accept the reward when their partner received a better reward for completing the same task. Surprisingly, the researchers also found that chimpanzees were more likely to refuse the high-value grape when the other chimpanzee received the low-value carrot than when the other chimpanzee also received a grape. The fact that chimpanzees were more likely to react negatively when their partner was not given an equal reward for their equal work than when they both received an equal reward reveals that chimpanzees react negatively to inequity even when they themselves are not the victims of the unequal distribution, and even when the inequity would actually benefit them. This demonstrates that at the very least, chimpanzees are able to detect inequity and can perceive that their partner received a less desirable reward, though the motivation behind their refusal remains unclear.

When capuchin monkeys and chimpanzees receive unequal rewards for performing an equal amount of work as their partner, the animals are likely to throw the reward back at the experimenter or refuse to continue participating in the experiment. Nonhuman primates’ negative reactions to inequity support an early evolutionary origin of inequity aversion. During the evolution of cooperation, it may have become crucial for individuals to compare their own efforts and rewards for those efforts with those of others. These studies provide evidence for the early evolutionary origins of our moral sense; non-human animals exhibit the ability to detect inequity, and further, this ability is not learned through explicit instruction.

Empathy, prosocial behaviors, and altruism

Compassionate emotions often prompt prosocial and altruistic behaviors, such as helping an agent achieve a goal or comforting someone in distress, and these behaviors emerge early in development. Sensitivity to the well-being of others, displayed by a dislike of harm or distress by emotional responses such as empathy, can also be observed very early in life.

Newborns exposed to the sound of another infant crying will display distressed reactions as soon as 18 to 72 hours after birth (Martin & Clark, 1982; Sagi & Hoffman, 1976; Simner, 1971). Newborns reacted more strongly to the sound of another infant’s cry than to other stimuli, including synthetic crying, non-human crying, and the sound of their own crying, which suggests that the distressed reactions displayed by infants were more than just a response to an unpleasant noise, but instead suggest an empathetic response. In Martin and Clark’s (1982) study, calm newborn infants cried in response to hearing the tape-recorded crying of other infants, yet calm infants who heard their own cry made almost no response, and calm infants also showed no response to the cries of a chimpanzee. They found that crying infants continued to cry when they heard the recording of another infant crying, yet when crying infants heard their own cry, they almost completely stopped crying. This suggests that infants’ emotional responses are caused by the distress of their peers. As soon as they are physically capable, infants begin to supplement these emotional responses with various prosocial behaviors, and even adapt their comforting behaviors to best suit the emotional needs of the other individual in distress. Types of prosocial behavior include comforting, defined as providing emotional support to others (Bischof-Kohler, 1992; Johnson, 1982; Zahn-Waxler et al, 1992), sharing, defined as giving food or objects to others (Hay et al., 1991); informing, defined as providing useful information for others (Dunn & Munn, 1986; Liszkowski et al., 2006); and instrumental helping, defined as acting on behalf of others’ goals (Rheingold, 1982; Warnenken et al. 2006; 2007).

Zahn-Waxler and colleagues (1992a) conducted longitudinal studies examining the development of empathy and related behaviors in young children between the ages of 14 and 36 months. Their experiments involved an adult experimenter pretending to be harmed (e.g. banging their knee and crying “ouch!”) while others observed and recorded infants and toddlers’ responses such as concern (e.g. sad looks or saying “I’m sorry”), prosocial behavior (e.g. hugs or asking “Are you ok?”), and empathetic responses such as personal distress. They found that by 14 months, infants exhibit both nonverbal and verbal concern for the adult experimenter, and that these empathetic behaviors increase significantly during the second year of life. By 24 months, nearly all children began to supplement their emotional responses with some form of helping behavior, and the qualities of such prosocial behaviors became more specified to suit the needs of the individual in distress during the second year of life. For example, younger infants reactions consisted mainly of physical attempts to soothe distress, while by 18 to 20 months, toddlers exhibited other forms of prosocial behaviors, such as verbal comfort, advice, sharing, and distraction.

Other studies have examined young children’s ability to cooperate with one another. A study by Brownell, Ramani, and Zerwas (2006) examined developments in children’s early cooperative abilities with peers. The researchers created a task in which 18 to 30-month-old children had to coordinate their actions with a peer in order to achieve a common goal. There were two versions of the task, though in both, each child had to perform the same behavior (pulling a handle) in order to activate a toy. In the first version of the task, each child pulled a different handle and the two handles needed to be pulled near the same time in order to activate the toy. In the second version, children needed to pull the handles sequentially in order to activate the toy (the correct actions were first demonstrated to the children by an adult experimenter in both versions). They found that 24 to 27-month-olds were able take another individual’s behavior into account and thus were able to cooperate successfully and help the other achieve their goal, and that by 30 months, toddlers were able to convey information to their peers about what to do. Their results suggest that the ability to cooperate with peers develops over the second and third years of life and develops in tandem with social understanding.

However, other studies have found that infants are able to coordinate their actions to help another individual achieve a goal even earlier in life when the infant’s social partner is an adult, not a peer. A study by Warneken and Tomasello (2006) investigated instrumental helping in 18-month-old infants. In the experiment, infants were presented with ten different situations in which an adult experiment had trouble attempting to achieve a goal. The difficulties with the goals fell into four different categories: (1) out-of-reach objects; (2) physical obstacles in the way of the goal (e.g. adult wants to put books in a cabinet but the doors are closed); (3) achieving a wrong result (e.g. adult wants to stack books but the book he places on top keeps slipping off the stack); or (4) attempting to achieve a goal through the wrong means (e.g. a spoon falls through a hole and the adult attempts to reach through the small hole to grab it, unaware of the flap on the other side of the box that can be lifted to retrieve the spoon). Their results showed that infants had a tendency to help the adults achieve their goals – infants handed the experimenter out-of-reach objects (but they would not do so if the experimenter had deliberately discarded the object); infants helped the experimenter stack the books (yet not if it appeared that the experimenter was deliberately misplacing the books); infants opened the door of the cabinet when the experimenters hands were full; and infants retrieved the spoon the experimenter could not access by lifting the flap on the box and handing it to him (but not if the experimenter had thrown that object inside the box purposefully).

The difference between infants ability to coordinate their actions with peers and their ability to coordinate their actions to help an adult achieve a goal is likely due to the fact that helping behaviors are more likely to occur if the behavioral or verbal cue for help is made more explicit. Previous studies on prosocial behaviors in young children have involved the experimenter giving an explicit cue, such as searching for an object, reaching for an object, failing to open something, or making a facial expression indicative of sadness or pain (Liszkowski et al., 2006; Dunfield & Kuhlmeier, 2010; Warneken & Tomasello, 2006; Dunfield, Kuhlmeier, O’Connell, & Kelley, 2011). These studies demonstrate that infants and toddlers have a tendency to help others achieve their goals in a variety of situations, even without being directly told to do so, which suggests that young children are naturally inclined to cooperate and to help others.

Empathy can be understood as an emotional response that prompts prosocial behaviors. de Waal has proposed that empathy is an evolved mechanism that promotes altruistic behavior (de Waal, 2008). In a review, Eisenberg and Miller (1987) noted that the ability to empathize often correlated with prosocial behaviors. Research has shown that infants as young as 18 months of age will help others to achieve their goals (such as opening cabinets) when distress cues are present (such as the adult looking confused) even in the absence of reward or reciprocation (Warneken & Tomasello, 2007). In fact, a similar study by Warneken and Tomasello found that the presence of material rewards actually diminished helping behaviors in 20-month-old infants, which suggests that young children are intrinsically motivated to help (Warneken & Tomasello, 2008).

Prosocial behaviors in non-human animals

Helping behaviors have also been observed in chimpanzees. The same study by Warneken and Tomasello (2007) that examined 18-month-old infants helping behaviors in response to ten different tasks also examined three young chimpanzees responses to similar helping tasks. The researchers sought to compare the performance of chimpanzees with that of infants on various tasks designed to assess prosocial behavior and attempted to determine whether instrumental helping is a uniquely human attribute. The same basic instrumental helping tasks were given (with some minor modifications) to three young chimpanzees. They found that the chimpanzees helped in some of the tasks. All three of the chimpanzees helped in the five tasks involving out of reach objects. Even when the chimpanzees had to go further out of their way to help the experimenter (when the object required more effort to retrieve), chimpanzees still helped. It may be the case that chimpanzees did not help the experimenter reliably in the other types of tasks (e.g. those involving physical obstacles, wrong results, or wrong means) because these tasks were more complicated than the out of reach ones. A human’s outstretched arm is quite a straightforward indication of their goal, and it seems likely that chimpanzees (and children) helped more in out of reach tasks because the goal and the need for help in such tasks were easier to understand. Overall, chimpanzees had a tendency to help the experimenter achieve his goal. Like the 18-month-old infants, the chimpanzees had a tendency to help even without receiving any benefit from the action (such as praise or rewards). The results demonstrate that our nearest primate relatives, chimpanzees, also behave prosocially and help others achieve their goals in the absence of rewards. The existence of instrumental helping toward goals in nonhuman primates supports the idea that altruistic behavior is a natural predisposition in social animals, and also lends support to the theory that such cooperative behaviors are the result of evolution intended to promote successful living among groups.

Warneken and Tomasello’s work demonstrates that chimpanzees will help humans achieve a goal, and a study by Horner et al. (2011) sought to determine whether chimpanzees would also behave prosocially toward other chimpanzees. In the experiments, two chimpanzees sat next to each other in separate cages, and one chimpanzee was given a bucket filled with thirty tokens (15 red and 15 green). When the chimpanzee selected the selfish (red) token, only the chimpanzee who had selected the token received a reward (food wrapped in paper). When the chimpanzee selected the prosocial (green) token, both chimpanzees were given a reward. Either choice the chimpanzee made resulted in the acting chimp receiving a reward, yet the results of the experiment found that chimpanzees overwhelmingly made the prosocial choice. When given the choice between helping only themselves or helping themselves and another chimpanzee, chimpanzees preferred to make the choice that would help both of them. This suggests that chimpanzees do care about the well-being of others, and also supports the idea that the drive to behave in an altruistic or prosocial manner when one can help someone else may be evolutionarily ancient. Additionally, there is abundant observational evidence that shows that, like humans, chimpanzees comfort others in distress (de Waal, 1996).


A moral system that evolved to facilitate cooperative behavior would require certain social evaluation skills – it would necessitate not only the ability to evaluate the actions and intentions of others, but also the ability to distinguish agents who are helpful (and would contribute to the success of the group) from agents who are unhelpful (and would detract from the well-being of the group). Beyond mere evaluative capacities, a system designed to promote cooperative behavior would require a sense of fairness and a sense of justice in regard to the distribution of vital resources and to the distribution of rewards and punishments. This means not only the ability to detect inequity or uncooperative behavior, but a propensity to respond negatively to inequity and uncooperative agents.

The literature reviewed in this paper make it clear that humans possess all of these abilities from a very young age and thus may have core knowledge of morality. By 12 months, infants evaluate goal-helping behavior as positive and goal-hindering behavior as negative. By about 8 months, infants are capable of evaluating agents based on their actions and use those evaluations to identify an agent as either helpful or unhelpful, and as early as 4 ½ months of age, infants prefer helpful agents to unhelpful or hindering ones. By the second year of life, infants have a complex sense of fairness; 12 to 18 month old infants prefer fair distributors to unfair distributors, and infer that others should likewise prefer a fair distributor, and 20-month-old infants expect rewards to be distributed according to effort (slackers should not receive the same reward as someone who worked). As young as 21 months of age, children have some sense of justice when it comes to evaluating the behavior of others, and prefer to distribute resources appropriately in accordance with this sense of justice – that is, toddlers prefer to take away resources from an antisocial or uncooperative individual and distribute resources to prosocial or cooperative individuals. Further, infants demonstrate empathy from a very early age. From birth, newborns display rudimentary signs of empathy by crying more at the distress of another baby. By the second year of life, infants and toddlers overwhelmingly tend to comfort others in distress and attempt to help others achieve a goal in any way they can.

We may infer that the displays of prosocial, altruistic, and cooperative behaviors detailed above stem from an innate source of motivation. It is unlikely that such behaviors are solely the result of socialization or learning, given that these behaviors occur very early in development, are also exhibited by our primate relatives, and decrease in the presence of external rewards. Indeed, research with non-human primates demonstrates that both human infants and chimpanzees appear intrinsically motivated to help others in need. Given that such altruistic tendencies emerge very early in life, it seems unlikely that their behaviors are a result of socialization and an internalization of an altruistic norm. It is more likely that children (and similar social animals) have a natural predisposition to behave altruistically. It is plausible that children begin as naïve or indiscriminate altruists, and that their behavior becomes more complex over time, in order to avoid being exploited by other individuals. The ability to identify “hinderers” or cheaters is an important facet of an innate moral mechanism evolved to promote altruistic behavior, as the absence of an ability to discriminate between those who may eventually reciprocate favors and those who most likely will never contribute would render altruistic behavior not evolutionarily viable. The early emergence of altruistic tendencies and the existence of altruistic tendencies in nonhuman primates (even in the absence of socialization practices) all give evidence for the claim that social animals have a predisposition for altruism, and that socialization is not the original source of altruistic tendencies.

The literature reviewed in this paper show that other social animals also exhibit abilities related to the moral sense. Chimpanzees behave prosocially toward conspecifics and will help a human experimenter reach out of reach objects. Capuchin monkeys and chimpanzees both react aversely to inequity, which shows that nonhuman primates have a sense of fairness. Additionally, chimpanzees are capable of social evaluation, base their evaluations on the actions of others, and use those evaluations to determine who would be a better cooperative partner.

The early emergence of social evaluation, a preference for fair distributions, a preference for cooperative and helpful individuals, empathy, and helping behavior make it seem quite plausible that such senses are innate and are not the product of socialization or learning, as these senses are displayed very early in life, before the child has really had much of a chance to acquire these senses through learning or instruction. It seems more likely that such senses are present from birth, and moral development builds upon these early-emerging social or moral senses. The fact that other social animals also exhibit a capacity for social evaluation, a sense of fairness, and prosocial behavior, makes it probable that such senses are evolutionarily ancient, and likely exist as a result of evolution favoring cooperation among individuals. Our moral sense guides us to do what is fair and often drives us to act kindly toward others. What we perceive to be our sense of morality stems from within, and leads us to get along with each other. As social animals, humans have long depended on one another for survival, and thus it seems likely that senses related to morality (fairness, social evaluation, empathy) are the product of evolutionary development favoring fair and cooperative interactions among individuals in order to facilitate successful group living.

Bates, E., Benigni, L., Bretherton, I., Camaioni, L., & Volterra, V. (1979) The emergence of symbols: Cognition and communication in infancy (pp. 111-128). New York: Academic Press
Szilvia Biro, Ã., Csibra, G., & Gergely, G. (2007). The role of behavioral cues in understanding goal-directed actions in infancy. From action to cognition, 164, 303.
Bischof-Köhler, D. (1992). The Development of Empathy in Infants. In M. E. Lamb & H. Keller (Eds.), Infant Development: Perspectives from German Speaking Countries, (pp. 245-273). Hillsdale , NJ : Erlbaum.
Boesch, C., & Boesch-Achermann, H. (2000) The Chimpanzees of the Tai Forest. (pp. 42-54). Oxford Univ. Press, Oxford
Bornell, C.A., & Carriger, M.S. (1990). Changes in cooperation and self-other differentiation during the second year. Child Development, 61, 1164–1174
Brosnan, S. F., Schiff, H. C., & De Waal, F. B. (2005). Tolerance for inequity may increase with social closeness in chimpanzees. Proceedings of the Royal Society B: Biological Sciences, 1560, 253-258.
Brosnan, S., Talbot, C., Ahlgren, M., Lambeth, S., & Schapiro, S. (2010). Mechanisms underlying responses to inequitable outcomes in chimpanzees, Pan troglodytes. Animal Behaviour, 79, 1229–1237.
Brosnan, S. F., & de Waal, F. B. (2003). Monkeys reject unequal pay. Nature, 425, 297-299.
Brownell, C. A., Ramani, G. B., & Zerwas, S. (2006). Becoming a Social Partner With Peers: Cooperation and Social Understanding in One‐and Two‐Year‐Olds. Child development, 77(4), 803-821.
Dunfield, K.A., & Kuhlmeier, V.A. (2010). Intention-mediated selective helping in infancy. Psychological Science, 21, 523–527
Dunfield, K., Kuhlmeier, V. A., O’Connell, L., & Kelley, E. (2011). Examining the diversity of prosocial behavior: Helping, sharing, and comforting in infancy. Infancy, 16(3), 227-247.
Dunn, J., & Munn, P. (1986). Siblings and the development of prosocial behaviour. International Journal of Behavioral Development, 9(3), 265-284.
Dupoux, E., & Jacob, P. (2007). Universal moral grammar: a critical appraisal. Trends in cognitive sciences, 11(9), 373-378.
Eckerman, C.O., Davis, C.C., & Didow, S.M. (1989). Toddlers’ emerging ways of achieving social coordinations with a peer. Child Development, 60, 440–453
Ecerkman, C.O., & Didow, S.M. (1989). Toddlers’ social coordinations: Changing responses to another’s invitation to play. Developmental Psychology, 25, 794–804
Eisenberg, N., & Miller, P. A. (1987). The relation of empathy to prosocial and related behaviors. Psychological bulletin, 101(1), 91.
Geraci, A. & Surian, L. (2011). The developmental roots of fairness: Infants’ reactions to equal and unequal distributions of resources. Developmental Science. 14(5), 1012‐1020
Greene, J. (2005). Emotion and cognition in moral judgment: Evidence from neuroimaging. In Neurobiology of human values (pp. 57-66). Springer Berlin Heidelberg.
Haidt, J. (2001). The emotional dog and its rational tail: a social intuitionist approach to moral judgement. Psychological Review, 108, 814–834.
Haidt, J. (2008). Morality. Perspectives on psychological science, 3(1), 65-72.
Hamlin, J. K. (2013). Moral Judgment and Action in Preverbal Infants and Toddlers Evidence for an Innate Moral Core. Current Directions in Psychological Science, 22(3), 186-193.
Hamlin, J. K., Wynn, K., & Bloom, P. (2007). Social evaluation by preverbal infants. Nature, 450(7169), 557-559.
Hamlin, J. K., Wynn, K., Bloom, P., & Mahajan, N. (2011). How infants and toddlers react to antisocial others. Proceedings of the national academy of sciences, 108(50), 19931-19936.
Hay, D. F., Caplan, M., Castle, J., & Stimson, C. A. (1991). Does sharing become increasingly” rational” in the second year of life?. Developmental Psychology, 27(6), 987.
Henderson, A. M., & Woodward, A. L. (2011). “Let’s work together”: What do infants understand about collaborative goals? Cognition, 121(1), 12-21.
Horner, V., Carter, J. D., Suchak, M., & de Waal, F. B. (2011). Spontaneous prosocial choice by chimpanzees. Proceedings of the National Academy of Sciences, 108(33), 13847-13851.
Johnson, D. B. (1982). Altruistic behavior and the development of the self in infants. Merrill-Palmer Quarterly (1982-), 379-388.
Killen, M., & Smetana, J. G. (2013). Morality: Origins and Development. To appear in M. Lamb & C. Garcia-Coll (Eds.), Handbook of child psychology, Vol. 3, Social and emotional development (7th edition; pp. 1-103). NY: Wiley/Blackwell Publishers.
Kuhlmeier, V., Wynn, K., & Bloom, P. (2003) Attribution of dispositional states by 12-month-olds. Psychological Science. 14(5), 402
Liszkowski, U., Carpenter, M., & Tomasello, M. (2008). Twelve-month-olds communicate helpfully and appropriately for knowledgeable and ignorant partners. Cognition, 108(3), 732-739.
Martin, G. B., & Clark, R. D. (1982). Distress crying in neonates: Species and peer specificity. Developmental psychology, 18(1), 3.
Melis, A. P., Hare, B., & Tomasello, M. (2006). Chimpanzees recruit the best collaborators. Science, 311(5765), 1297-1300.
Premack, D. (1990). The infant’s theory of self-propelled objects. Cognition,36(1), 1-16.
Premack, A.J., & Premack, D. (1997). Infants attribute value ± to the goal-directed actions of self-propelled objects. Journal of Cognitive Neuroscience, 9(6), 848-856.
Range, F., Horn, L., Viranyi, Z., & Huber, L. (2009). The absence of reward induces inequity aversion in dogs. Proceedings of the National Academy of Sciences, 106(1), 340-345.
Rheingold, H. L. (1982). Little children’s participation in the work of adults, a nascent prosocial behavior. Child Development, 114-125.
Ross, H.S., & Lollis, S.P. (1987) Communication within infant social games. Developmental Psychology, 23, 241–248
Sagi, A., & Hoffman, M. L. (1976). Empathic distress in the newborn. Developmental Psychology, 12(2), 175.
Simner, M. L. (1971). Newborn’s response to the cry of another infant. Developmental Psychology, 5(1), 136.
Sloane, S., Baillargeon, R., & Premack, D. (2012). Do infants have a sense of fairness?. Psychological science, 23(2), 196-204.
Spinrad, T. L., Eisenberg, N., Cumberland, A., Fabes, R. A., Valiente, C., Shepard, S. A., . . .Guthrie, I. K. (2006). Relation of emotion-related regulation to children’s social competence: a longitudinal study. Emotion, 6(3), 498.
Turiel, E. (2006). The development of morality. In N. Eisenberg, W. Damon & R. M. Lerner (Eds.), Handbook of child psychology: Social, emotional, and personality development (pp. 789- 857). Hoboken, NJ: Wiley.
de Waal, F. B. (1996). Good natured (No. 87). Harvard University Press.
de Waal, F. B. (2008). Putting the altruism back into altruism: The evolution of empathy. Annual Review of Psychology, 59, 279-300.
Warneken, F., Chen, F., & Tomasello, M. (2006). Cooperative activities in young children and chimpanzees. Child development, 77(3), 640-663.
Warneken, F., Hare, B., Melis, A. P., Hanus, D., & Tomasello, M. (2007). Spontaneous altruism by chimpanzees and young children. PLoS biology, 5(7), e184.
Warneken, F., & Tomasello, M. (2006). Altruistic helping in human infants and young chimpanzees. Science, 311(5765), 1301-1303.
Warneken, F., & Tomasello, M. (2007). Helping and cooperation at 14 months of age. Infancy, 11(3), 271-294.
Warneken, F., & Tomasello, M. (2008). Extrinsic rewards undermine altruistic tendencies in 20-month-olds. Developmental psychology, 44(6), 1785.
Warneken, F. & Tomasello, M. (2009). The Roots of Human Altruism, British Journal of Psychology. 100, 455–471
Watts, D., & Mitani, J. (2002) The Diversity of Chimpanzees. In C. Boesch, G. Hohmann, L. Marchant (Eds.), pp. 244–255, Cambridge University Press, Cambridge
Woodward, A. L. (2009). Infants’ grasp of others’ intentions. Current directions in psychological science, 18(1), 53-57.
Zahn-Waxler, C., Radke-Yarrow, M., Wagner, E., & Chapman, M. (1992). Development of concern for others. Developmental psychology, 28(1), 126.
Zahn-Waxler, C., & Robinson, J. (1995). Empathy and guilt: Early origins of feelings of responsibility. In K. Fischer & J. Tangey (Eds.), Self-Conscious Emotions: Shame, Guilt, Embarrassment and Pride, (pp 143-173). New York: Guilford.