In the summer of 1954, twenty-two ordinary boys—white, middle-class, Protestant, well-adjusted eleven- and twelve-year-olds with no history of behavioral trouble—arrived by bus at a remote Boy Scout camp in Robbers Cave State Park, Oklahoma. They believed they were simply at summer camp. None knew they were subjects in a psychological experiment. The researchers, led by Muzafer Sherif and posing as counselors, janitors, and caretakers, had carefully matched the boys for talent and temperament, then randomly split them into two groups that never initially knew of each other’s existence.
Within days the boys formed tight bonds. They hiked, swam, and worked together. They named themselves the Eagles and the Rattlers, stenciled the names on shirts and flags, and developed hierarchies, norms, and fierce loyalty. Near the end of the first week they discovered the other group. “Us and them” language appeared almost immediately. Both sides pressed the staff to let them challenge the outsiders—before any formal competition had begun.
Then the experimenters introduced staged scarcity and zero-sum contests: baseball, tug-of-war, a tournament with a single trophy, medals, and multi-bladed knives for the winners only. No consolation prizes. Situations were engineered so one group’s gain meant the other’s loss—delayed picnics where the latecomers found the food already eaten. Confidence turned into taunts, then into raids. Flags were burned. Cabins were ransacked. Beds overturned. Property stolen. Staff had to physically separate the boys. During a cooling-off period each side described its own group in glowing terms and the other in harshly negative ones. These were not street toughs. They were the same well-adjusted boys who had arrived days earlier.
Sherif’s realistic conflict theory emerged from this: intergroup hostility arises when groups compete for limited resources. In-group favoritism and out-group hostility become standardized. Contact alone did not heal it. Shared meals, movies, and fireworks failed. Only superordinate goals—problems neither group could solve alone—worked. A sabotaged water supply required joint effort to clear a stuffed faucet. A desired movie required both groups to pool scarce cash. A truck stuck in a rut carrying food for both camps demanded cooperation. Hostility eased. By the end the boys mixed freely at meals, chose cross-group friendships in roughly two-thirds of cases, and asked to ride home together on one bus.
The study remains one of the most cited in social psychology. Roger Brown called it “the most successful field experiment ever conducted on intergroup conflict.” Yet its power lies less in the tidy three-stage narrative than in the uncomfortable recognition that the techniques used to produce the conflict are neither rare nor confined to a 1950s Oklahoma camp.
The Techniques of Polarization
Divide people who have no prior animus into identifiable groups. Give each group a name, symbols, and shared activities that build internal loyalty and hierarchy. Introduce competition over scarce, exclusive rewards—status, resources, recognition, victory. Frame outcomes as zero-sum so one side’s success is the other’s humiliation. Amplify the resulting friction through selective attention and narrative. When hostility appears, present it as spontaneous and inevitable rather than engineered. These are the core moves of Robbers Cave.
Modern versions rarely require a remote park and counselors with clipboards. They operate at scale through media, institutions, political incentives, and digital platforms. Group boundaries are drawn along partisan, racial, regional, religious, educational, or cultural lines. Scarcity is framed around jobs, housing, status, moral prestige, or institutional power. Algorithms and editorial choices reward the most vivid expressions of out-group threat. Identity becomes a team jersey. Disagreement becomes disloyalty. The other side is not merely wrong; it is a danger to the in-group’s survival or dignity.
The result is measurable polarization. People rate their own side more favorably and the opposing side more harshly, just as the Eagles and Rattlers did. Trust erodes. Cooperation on shared problems becomes harder. Ordinary citizens who might otherwise find common ground on concrete issues—safe streets, functioning schools, economic opportunity—are recruited into tribal contests where the prize is moral superiority or electoral dominance and the cost is mutual contempt.
Critiques of the original experiment sharpen rather than dismiss the parallel. The sample was tiny, homogeneous, and all male. The rivalry was artificial and short-lived. Staff quietly influenced tournament outcomes and staged the later crises that produced reconciliation. Gina Perry’s archival work on Sherif’s abandoned 1953 pilot study at Middle Grove revealed boys who detected the adult manipulation and redirected their resentment at the researchers instead of each other, forcing the study’s cancellation. Social identity theory demonstrated that mere categorization into groups—without material competition—can produce bias. Culture matters: other studies found children from more cooperative societies less prone to the same hostility. Prior friendships and cooperative leadership can blunt competitive conflict, as Tyerman and Spencer observed with English scouts.
These limitations do not erase the mechanism. They show that hostility is not automatic or universal. It is more readily produced under certain conditions: unfamiliar groups, engineered scarcity, absence of higher shared purpose, and active amplification by those who benefit from the conflict. When those conditions are deliberately or carelessly replicated at societal scale, polarization follows.
Shared Goals as the Antidote—and Why They Are Hard
Sherif’s third phase remains the most practical finding. Superordinate goals that require genuine interdependence reduced hostility where contact and communication failed. The same principle has been applied deliberately: Elliot Aronson’s jigsaw classroom made multi-ethnic students interdependent for academic success and reduced prejudice. Organizations that replace departmental competition for fixed budgets with shared metrics tied to collective outcomes can reduce internal turf wars. Communities facing genuine external threats—natural disasters, infrastructure failures, public-health crises—often discover temporary cooperation across old lines.
Yet the modern environment makes such goals difficult to sustain. Political and media incentives frequently reward the opposite: framing every issue as existential competition, treating compromise as betrayal, and monetizing outrage. When institutions themselves become arenas of group contest rather than neutral ground for solving shared problems, the water faucet stays clogged and the truck remains stuck. Polarization then becomes self-reinforcing. Each side’s hostility is offered as proof that the other side is the real threat, justifying further escalation.
The boys at Robbers Cave left on better terms than they arrived. That outcome was not inevitable. It required the deliberate introduction of problems larger than either group’s immediate interests. Societies that treat polarization as natural or permanent forfeit the same possibility. The experiment’s enduring warning is not that conflict is inevitable among groups. It is that conflict can be manufactured among ordinary people with ordinary loyalties—and that the same people are capable of cooperation when the incentives and the stakes are redesigned around what they must accomplish together rather than what they must defeat in each other.
0 Comments