Us vs. them: as simple as ABC?

If you, like me, are discouraged by our human instinct to focus on superficial differences rather than fundamental similarities, a new psychology study will interest you. But I'm not sure whether it's good news or bad news.

Turns out we don't even need markers like skin color, ethnicity, or hair style to divide ourselves from other people, according to researchers at Yale and elsewhere. We don't even need other people—just "80 lines of computer code."

Using "game-theoretical agent-based models," the scientists wrote a program that would prompt virtual people to form friendships based on two simple principles: reciprocity and transitivity.

Reciprocity is straightforward: if you like me, then I like you. If not, then not. Transitivity extends that notion to friends of friends and enemies of friends.

"Imagine a triad of people—A, B, and C," the authors write. "If A and B are friends, and A (dis)likes C, then B should also (dis)like C."

Led by David Rand of Yale and Kurt Gray of the University of North Carolina, the reseearchers "created a computer simulation that assigned numerical values to how much each pair of individuals in a population 'liked' each other," a Yale news release explains.

The results: "Pairs of friends and enemies would form at random, but not in larger groups. However, groups formed robustly once the simulation incorporated 'friends of friends' effects." The paper will be published in the journal Psychological Science.

The computer simulation has an advantage over real-life observation, where “it is hard to tease out root causes of group formation because observable differences such as race or language may arise because people already live in groups,” Rand argues. “What we show is that you can strip away all those factors, and groups will still emerge.”