The evolutionary reasons behind who we choose as friends

A new study explores the evolutionary roots of friendship, suggesting we choose same-sex friends based on traits that would have made them valuable cooperative partners in ancestral environments—not just how useful they are to our current goals.

People may choose same-sex friends partly based on traits that would have made someone a valuable cooperative partner in ancestral environments, according to research published in Evolution & Human Behavior . Friendship is often explained through present-day usefulness. According to this view, people are drawn to potential friends who help them pursue current goals, whether by offering kindness, intelligence, support, shared interests , or practical benefits.

But this explanation leaves open whether people’s friend preferences are shaped only by what is useful now, or whether some preferences reflect older social pressures from environments in which choosing the right cooperative partners mattered for survival and reproduction. Adar B. Eisenbruch and colleagues examined the possibility that same-sex friendship preferences partly reflect evolved cooperative partner choice. Because ancestral humans depended on long-term cooperation for tasks such as foraging, childcare, social support, and conflict management, people may have come to value friends who seemed able and willing to create and share benefits.

From this perspective, traits such as prosociality, social status, physical attractiveness, and perceived survival-related skill may be appealing not only because they are useful for people’s current goals, but also because they historically signaled a valuable cooperative partner. In Study 1, the researchers recruited 156 students from a public university in the United States for a lab-based interaction study. Participants were paired with an unfamiliar same-sex partner and completed a structured 30-minute face-to-face conversation.

Afterward, they privately rated how desirable their partner was as a potential friend and evaluated the partner on traits thought to matter for friendship, including ancestral productivity, prosociality, dominance, social status, attractiveness to the other sex, nurturance, and academic success. The researchers also separately measured participants’ physical strength using bicep circumference, grip strength, and chest strength. As a behavioral indicator of friendship interest, participants could choose whether to share their email address with their partner, but contact information was exchanged only if both members of the pair agreed.

In Study 2, 444 U.S. participants recruited through Amazon Mechanical Turk first described their most important current goal. They then rated photographs of unfamiliar same-sex faces based on how much they would want to be friends with each person. Afterward, they judged how helpful or harmful different traits in a new friend would be for achieving their stated goal.

A separate group of raters evaluated the same faces on traits such as Big Five personality characteristics, attractiveness, intelligence, trustworthiness, aggressiveness, competence, and ancestral productivity. This allowed the researchers to compare whether the traits that made faces seem desirable as friends were simply the same traits participants saw as useful for their current goals. In Study 1, participants were more interested in friendship with partners who appeared higher in ancestral productivity, prosociality, dominance, social status, attractiveness, and nurturance.

When the researchers considered the traits together, attractiveness, prosociality, and dominance emerged as the strongest unique predictors of friend desirability. Men were also more likely than women to offer their contact information, even though they did not rate their partners as more desirable overall. Study 2 showed that friend preferences were not fully explained by current goal pursuit.

In particular, attractiveness and ancestral productivity predicted friend desirability more strongly than would be expected based on how relevant those traits were to participants’ goals. This suggests that people’s same-sex friend preferences may be shaped not only by present-day usefulness, but also by preferences for traits that historically signaled a valuable cooperative partner. The