Project 9

Human Social Preferences

The ontogney of altruistic social preferences
Children palying a game on a tablet. Credit: WikiLouise / Wikimedia

Principal Investigators

ASU Collaborators

Project 9 fills gaps in our understanding of how child cognitive development interacts with culture (beliefs, norms, and values) and socialization to generate cross-cultural variation in cooperation in several small-scale human societies. The ontogeny of culturally specific patterns of cooperation connects to Project 10’s focus.

With clear links to our proposed study of inter-population variation in human and ape life-history attributes (Project 8), this project will enhance our understanding how cognitive development interacts with culture (beliefs, norms, and values) and socialization to generate cross-cultural variation in human cooperation. It will contribute to our understanding of the relationship between prosocial behavior that motivates cooperation in children and the role of culture in shaping this process.

Detailed Project Description

Humans differ from all other organisms in our capacity for large-scale cooperation that is not based on kinship or reciprocity. There is, however, substantial variation in the nature of cooperation across societies. It seems likely that these differences are maintained by learned cultural beliefs that are transmitted from one individual to another within societies. These sets of beliefs, norms, and values may have been shaped by cultural evolution to maintain group-level cooperation.

Although cross-cultural differences in the cooperative behavior of adults have been documented, there is little systematic information about the developmental trajectory that underlies this variation. How and when does the cooperative behavior of children begin to resemble the behavior of adults from their own societies? Although developmental psychologists have documented important changes in empathy and helpful behavior as children mature, relatively little of this work has been done in a cross-cultural context.

To address this gap, we previously examined the ontogeny of culturally specific patterns of generosity among children and adolescents in five small-scale societies and one urban western population. In one of the experimental tasks, the Costly Sharing Game, subjects were presented with a choice between two options. One option provided a reward to the subject and an identical reward to another child, and the other option provided two rewards to the subject and nothing to the other child. Thus, generosity was costly. There was a decline in children’s likelihood of choosing the generous option as they matured from 3 to about 8 years of age in all populations, suggesting that all children became progressively more concerned with rewards for themselves as they got older. Cultural differences emerged as children reached middle childhood (7–9 years), and older children tracked toward the behavior of adults from their own societies in the same task. In contrast, we found no evidence for variation across societies in a task in which children could provide benefits to others at no cost to themselves. Based on these findings, we hypothesize that middle childhood is a particularly important period for the acquisition of culturally specific values, norms, and beliefs that influence cooperation. Moreover, cultural variation is most pronounced in situations that require personally costly forms of cooperation.

The primary goals of the proposed cross-cultural project are to critically evaluate this hypothesis by (1) documenting the developmental trajectories of a wider range of altruistic social preferences, (2) assessing psychological biases that may underlie decision making and expression of altruistic social preferences; and (3) examining the development of children’s judgment about resource allocation decisions. In our previous experimental work, we examined generosity. Here, we will extend this work to assess fairness, inequity aversion, and third-party punishment. In addition, we will examine some of the psychological traits that may influence prosocial behavior. Time preferences, the willingness to trade immediate payoffs for future gains, vary across societies. Within societies, individual variation in time preferences is associated with prosocial behavior in adults and preschoolers. Finally, we will examine children’s reasoning about decision making in these contexts, by eliciting third-party judgments of the behavior of others in hypothetical scenarios.


We will use a set of experimental tasks have been developed by behavioral economists and developmental psychologists to assess social preferences and psychological biases. By using standardized tasks, we will be to compare results derived from our study with results derived from previous work.


In the Prosocial Game and the Costly Sharing Game, a subject chooses one of two options that provide payoffs to herself and her partner (see table). We will also conduct a standard Dictator Game in which the subject is given an endowment (e.g., 10 crackers), and can allocate any amount of the endowment to a recipient.

Inequity Aversion

The Envy game assesses actors’ preferences for options that benefit others more than they benefit themselves (see table). In the Inequity Game, the experiment presents an offer that provides A rewards to the actor and R rewards to the recipient. The actor is able to decide whether the offer will be accepted or rejected. If the offer is accepted, rewards are distributed. If the offer is rejected, neither child gets any rewards. Actors are presented with offers that are even; advantageous to themselves (A=4, R=1), or disadvantageous to themselves (A=1, R=4). In both the Envy Game and the Inequity Game, children’s decisions to choose or accept lower payoffs are a measure of their responses to unfair offers. We will also conduct a Mini Ultimatum Game. In this game the recipient can accept or reject the subject’s choice. If the recipient accepts the subject’s choice, each child gets the designated number of rewards. However, if the recipient rejects the subject’s choice, neither child receives any rewards. Recipients have little material incentive to reject any nonzero offer, so rejections constitute a form of costly retaliation for undesirable offers.

Punitive Sentiments

For further information about punitive sentiments, we will also conduct a Third Party Punishment Game. This game combines a standard Dictator Game with a punishment phase in which an observer is given an endowment, and can use some of the endowment to impose a penalty on the actor. In this game, punishment is altruistic because the observer gives up resources in order to impose costs on the actor.

Time Preferences

We will use a version of the classic “marshmallow game.” In this game, children are told that they can have one reward (e.g., marshmallow) now, or wait until the experimenter returns and get a larger number of rewards. Children’s ability to resist eating the marshmallow, or the latency to consumption, provides a measure of their ability to delay gratification. This measure is a reliable correlate of school performance and social preferences in western societies.

Judgments about resource allocation

We will evaluate children’s responses to hypothetical scenarios about resource allocations in which we manipulate the cost to the actor, evenness of the distribution, source of the endowment (earned/unearned), and target of the allocation (ingroup/outgroup).

Study Sites

We will work in collaboration with researchers who have established long-term projects in small-scale societies. These researchers have deep knowledge about subsistence practices, kinship networks, and cultural values, practices, and beliefs in the societies in which they work, and this information will enrich our analyses. Working with researchers at established sites also provides many logistical advantages because we can share equipment, facilities, and personnel. Sites will be chosen to provide diversity in the prosocial behavior of adults because we are interested in how children acquire culturally specific values and beliefs that influence altruistic behavior. We will continue to collaborate with researchers at some of the sites where we have worked before (Joseph Henrich in Fiji, Brooke Scelza in Namibia) in order to extend and replicate previous findings, and have begun discussions with ethnographers working in several other small-scale societies (Tanya Broesch in Vanuatu, Sarah Mathew in Kenya, and Polly Weisner in Botswana).