The Robbers Cave experiment was a 1954 field study in social psychology that demonstrated how quickly arbitrary group divisions can produce genuine hostility between people, and how shared challenges requiring cooperation can dissolve that hostility. Conducted by Muzafer Sherif and his colleagues at Robbers Cave State Park in southeastern Oklahoma, the study used twenty-two boys at a summer camp as unwitting participants in a staged conflict. The results became one of the most cited demonstrations of how intergroup prejudice forms and, more usefully, how it can be reversed.
The Setup at Robbers Cave State Park
Sherif’s team recruited twenty-two white, middle-class boys, all roughly eleven or twelve years old, from the Oklahoma City area. The boys were carefully screened to be psychologically well-adjusted, with no prior behavioral problems or unusual family stressors. None of them knew each other before arriving at the camp. The researchers split them into two groups of eleven and kept the groups unaware of each other’s existence for the first week.
The study was designed in three distinct phases. During the first phase, each group formed its own identity through typical camp activities: hiking, swimming, cooking meals. One group named themselves the Rattlers; the other chose the Eagles. Each group developed its own internal hierarchy, norms, and slang. The researchers, posing as camp counselors, watched and documented everything. By the end of the first week, each group had a strong sense of “we,” complete with loyalty norms and mild peer pressure to conform.
How Conflict Took Hold
The second phase began when the groups discovered each other. The researchers arranged a series of competitive activities, including baseball games, tug-of-war, and a treasure hunt, with prizes going to the winning group. The transition from friendly competition to real animosity happened remarkably fast. Within days, the boys were calling each other names, refusing to eat in the same dining hall, and raiding each other’s cabins. The Eagles burned the Rattlers’ flag. The Rattlers retaliated by trashing the Eagles’ cabin.
What made the hostility notable was its depth. These were not casual insults tossed around during a game. The boys developed genuine negative stereotypes about the other group, rating out-group members as untrustworthy, dishonest, and inferior. In-group solidarity tightened: boys who had been peripheral figures in their own group became more loyal and more willing to follow group norms when facing a common rival. The researchers documented this shift using rating scales and direct observation, finding that unfavorable characterizations of the out-group became the standard view rather than the exception.
The speed and intensity of the conflict surprised even the researchers. These boys had no history of enmity, no cultural grievances, no differences in race or class. The only thing separating them was an arbitrary group assignment and a competitive structure. That was enough to produce hostility that looked, in miniature, like the kind of prejudice seen in the wider world.
What Superordinate Goals Actually Changed
The third phase tested Sherif’s central hypothesis: that intergroup conflict could be reduced not by mere contact between groups, but by introducing problems that neither group could solve alone. These shared challenges are called superordinate goals. The researchers staged a series of them. The camp’s water supply was “sabotaged” (by the researchers), forcing both groups to work together to find and fix the problem. A truck needed to transport food for both groups “broke down,” and the boys had to pull it with a rope, a task that required everyone’s effort. A movie the boys wanted to watch had to be “rented” with pooled funds from both groups.
Early cooperative encounters did not immediately erase the hostility. The first joint task still featured name-calling and blame-shifting. But as the superordinate goals accumulated, attitudes shifted. Boys who had been singled out as enemies became partners. By the end of the cooperation phase, the proportion of favorable ratings that boys gave to members of the other group rose from about 15 percent to over 86 percent.1Academia.edu. Intergroup Conflict and Cooperation: The Robbers Cave Experiment The boys began sitting together at meals, choosing friends from the other group, and on the bus ride home, the Rattlers used prize money they had won to buy malts for the Eagles.
The critical insight was not just that cooperation helped, but that a single cooperative event was not enough. Repeated experiences of working together toward goals that mattered to both groups were necessary before the deep attitudinal shift occurred. Contact alone, without interdependence, had no effect. When the researchers had earlier tried simply bringing the groups together for meals or movies without a shared task, the contact actually increased friction rather than reducing it.
The Failed Experiment Nobody Talks About
Robbers Cave was not Sherif’s first attempt at this kind of study. In 1953, a year earlier, he and his team tried a nearly identical experiment at a camp near Middle Grove, New York. That study fell apart. The boys figured out they were being manipulated. Some became suspicious of the “counselors” who kept steering activities in oddly specific directions. The intergroup hostility phase got out of hand in ways the researchers could not manage, and the study was abandoned before the cooperation phase could be tested.
This failed predecessor matters for understanding Robbers Cave, because it reveals the degree to which the 1954 study’s smooth arc depended on the researchers’ active management. The boys at Robbers Cave did not simply drift into conflict and then drift out of it. The researchers designed situations to provoke hostility and then designed situations to encourage reconciliation. The question of how much the outcomes reflect natural group dynamics versus the researchers’ skill at stage management has been debated ever since.
Where Later Studies Agreed and Disagreed
Attempts to replicate the Robbers Cave findings have produced a mixed picture. One notable replication effort tested Sherif’s core hypotheses using a similar summer camp design. The results diverged from the original in an interesting way: in-group solidarity did not increase during the camp as Sherif’s framework predicted. In fact, it decreased. However, the study did find support for Sherif’s ideas about how individuals respond to appeals for intergroup cooperation, and it confirmed that differences in group composition and motivation shaped how intergroup relations played out.2Sage Journals (Small Group Behavior). A Critical Test of the Sherifs’ Robber’s Cave Experiments
The partial failure of replication points to something important. Sherif’s theory predicted a tidy sequence: groups form, competition breeds hostility, cooperation breeds harmony. Real groups do not always follow that script. Whether in-group cohesion tightens under competition depends on how the group is composed, what the members care about, and how the competitive situation is structured. The superordinate goals finding has held up better than the hostility-formation finding, which is arguably the more useful result. Knowing how to reduce conflict matters more than knowing how easily it starts.
Ethical Problems That Would Stop the Study Today
By modern research ethics standards, the Robbers Cave experiment could never be conducted. The boys and their parents did not give informed consent to participate in a psychological experiment. The parents were told their sons were going to a summer camp, which was technically true, but the deliberate manipulation of the boys’ social environment was concealed. The researchers intentionally provoked anger, fear, and aggression in children, then watched to see what happened.
Beyond consent, the study involved deception at every level. The “counselors” were researchers. The broken water supply was staged. The truck breakdown was fake. The boys had no idea that their friendships, rivalries, and emotional reactions were being systematically documented. When the conflict phase escalated to cabin raids and physical confrontations, the researchers allowed it to continue because the escalation was the point of the study. Intervening would have compromised the data.
Critics have also raised concerns about the researchers’ role in shaping outcomes. Gina Perry, a science historian who examined Sherif’s archives, found evidence that the researchers were more involved in stoking conflict than the published reports acknowledged. In some cases, staff members actively encouraged competitive hostility when it was not developing fast enough on its own. This does not necessarily invalidate the findings about superordinate goals, but it complicates the claim that intergroup hostility emerged “naturally” from competition alone.
Why Simple Contact Is Not Enough
One of the most durable takeaways from Robbers Cave is the distinction between contact and cooperation. In the years before Sherif’s work, the dominant idea in prejudice reduction was the “contact hypothesis,” which held that prejudice results from unfamiliarity, so bringing groups together should reduce it. Sherif’s experiment showed this was incomplete. When the Rattlers and Eagles were brought together without a shared task, they used the opportunity to hurl insults and throw food. Proximity without purpose made things worse.
This finding reshaped how social scientists thought about desegregation, workplace diversity, and international peace-building. It suggested that policies bringing hostile groups into the same physical space, such as shared schools or integrated neighborhoods, would not automatically produce harmony. The groups needed a reason to depend on each other. In educational settings, this insight eventually influenced the development of cooperative learning techniques, where students from different backgrounds are placed in small groups with shared academic goals that require everyone’s contribution.
The principle has shown up in conflict resolution efforts far beyond summer camps. Programs that bring together youth from opposing sides of political or ethnic conflicts often structure the experience around joint projects rather than simple dialogue. The logic traces directly back to Sherif’s superordinate goals: shared work toward a common outcome creates bonds that conversation alone does not.
What Neuroscience Has Added to the Picture
Decades after Robbers Cave, neuroscience researchers began investigating the brain processes underlying the group behaviors Sherif observed. Brain imaging studies have explored how people respond differently to in-group and out-group members at a neurological level, examining the psychological and biological factors behind group formation, cooperation, and conflict.3PubMed. The Neuroscience of Intergroup Relations: An Integrative Review This research has found that the brain processes faces and actions of in-group members differently from those of out-group members, often with more empathy and more nuanced emotional responses directed toward people perceived as “us.”
These findings help explain why the boys at Robbers Cave shifted their attitudes so rapidly. The brain’s categorization of “us” and “them” is flexible. Once the superordinate goals restructured the situation so that “us” included everyone pulling the rope or fixing the water supply, the neurological responses associated with in-group warmth extended to former rivals. The malleability of these brain responses is, in some ways, the most encouraging part of the story. Group boundaries feel rock-solid when you are inside them, but they can be redrawn surprisingly quickly when circumstances change.
How the Experiment Shaped Real-World Policy
Sherif’s work arrived at a moment when the United States was wrestling with desegregation, and it offered a framework that went beyond moral appeals. The idea that groups in conflict need shared, urgent, practical goals to overcome hostility filtered into organizational psychology, military integration planning, and international relations theory. The concept of superordinate goals became a standard tool in conflict mediation: find something both sides need that neither can achieve alone, and use it as the basis for cooperation.
In corporate settings, the principle shows up in cross-functional team design. When departments that typically compete for resources, such as sales and engineering, are given a shared objective with shared accountability, the inter-departmental friction tends to decrease. The mechanism is the same one Sherif identified: interdependence shifts the boundary of “my group” to include people who were previously outsiders.
International diplomacy has drawn on similar logic. Joint scientific projects, shared infrastructure development, and cooperative disaster response have all been used as tools to build trust between nations with adversarial histories. The results are not always as dramatic as the shift from 15 percent to 86 percent favorable ratings, but the underlying principle, that working together on something real beats talking about getting along, has proved remarkably robust across contexts.
Children, Gender, and Group Bias Beyond Robbers Cave
The Robbers Cave experiment used only boys, which raises the question of whether the same dynamics apply across genders and age groups. Research on children’s intergroup behavior has expanded significantly since 1954, exploring how group biases form along lines of gender, ethnicity, and other social categories.4British Journal of Developmental Psychology. Children’s intergroup gender bias and self‐perceived same‐ and other‐gender similarity Studies on children’s gender-based group bias have found that kids develop strong in-group preferences early, often rating same-gender peers more favorably and perceiving themselves as more similar to same-gender children than to other-gender children.
These findings suggest that the group dynamics Sherif observed were not unique to his particular sample of boys in Oklahoma. The tendency to favor an in-group and develop negative views of an out-group appears across different kinds of group boundaries, including ones that are not experimentally created but socially inherited. What Sherif captured in a compressed two-week window, using arbitrary group labels and competitive games, mirrors a process that unfolds more gradually in everyday social development. The same malleability also appears: children’s intergroup biases can be influenced by the structure of their social environment, including whether they experience cooperative or competitive interactions with members of other groups.
What the Experiment Cannot Tell You
For all its influence, Robbers Cave has significant limits. The participants were a remarkably homogeneous group: all white, all male, all middle-class, all from the same city, all roughly the same age. The conflict between the Rattlers and Eagles had no history, no economic stakes, and no structural power differences. Real-world intergroup conflicts almost always involve at least some of those complicating factors. A study that creates hostility from scratch in two weeks cannot fully model a conflict rooted in decades of inequality or historical trauma.
The superordinate goals principle also has boundaries. It works best when both groups have roughly equal status and roughly equal ability to contribute to the shared task. If one group is significantly more powerful or better resourced, cooperation can feel like coercion to the weaker group, and the resulting “harmony” may be more about submission than genuine attitude change. Sherif’s boys were evenly matched by design. The world rarely is.
There is also the question of durability. The boys’ attitudes shifted dramatically over the course of a few days of cooperation, but the study ended when camp ended. Whether those positive feelings persisted once the boys returned to their normal lives, with no further interdependence and no further contact, was never tested. Short-term attitude change in a controlled environment is not the same as lasting prejudice reduction in the real world, and assuming it translates directly has led to some overly optimistic applications of Sherif’s ideas in settings where a weekend workshop is expected to undo years of entrenched bias.