If some sociologists are correct, we are either more discerning about who our friends are, or more isolated than ever. One 2011 Cornell University study found that the median number of “close friends” in the United States over the past 30 years has dropped from three to around two. Other researchers have claimed that Facebook can make us lonely, despite its founder’s vision that our ballooning friend counts inspire better relationships.
But before we decry the state of our modern emotional life, hopelessly tagged to a generational malaise and virtual interaction, here’s another set of observations to consider: Today, a team of researchers at Oxford and Aalto University in Finland published a study looking at how 24 graduating British high school students’ relationships changed over the course of 18 months. When the researchers tracked the phone records and surveyed friendships of those who entered the workforce, stayed home, or left for university, they came to the surprising conclusion that not much had changed at all.
That’s not to say that the friends, or social networks, remained the same. While the identities of friends within the subjects’ “ego networks” shifted, researchers found that the “social signatures,” or the time and care devoted to members of that network did not. A person could have three close friends, say, then move to another city and set up a similar network of closeness with three entirely new people. And, no matter the circumstance, the graduating students held that pattern.
The participants devoted a median 25% of their time on the phone to one top-ranked friend and roughly 48% of their overall time to their three top-ranked friends together, according to the study, which was published in the Proceedings of the National Academy of Sciences. The researchers asked the students to assess the closeness of their relationships and found this self-reporting correlated with the phone data.
“People tend to structure their social networks fairly similarly, regardless of circumstances, and regardless of who’s in those networks,” explains lead scientist Jari Saramaki. “The signature patterns do not change that much, although many of the friends get replaced by newcomers.”
What this tells us, Saramaki says, is that our capacity to have close relationships with a large number of people is limited. British anthropologist Robin Dunbar might’ve been onto something when he suggested that a human can only maintain 150 relationships at a time. The PNAS paper also suggests that the number of close friends we’re able to maintain simply has to do with a fixed amount of emotional capital, or simply the amount of devotion we’re able to spread around.
The team’s observations could be heartening, in a sense, if you take them one step further. Though the researchers didn’t look at what happened past the 18 months of observation or at data from another age group, the findings suggest that the structure of our close friendships will persist, despite life changes, or perhaps even significant changes in technology–though that remains to be proven.
Saramaki acknowledges that his study didn’t measure online communication and to what degree this affected close friendships. Still, the persistence of the social signatures over periods of big life changes in his study seemed to suggest that our emotional capital, rather than our technologies or geographic location, might dictate our social habits.
“This is again, just my own thinking, but I find it funny that all new communication technologies are extremely unpredictable. Mobile phones, they were designed for businessmen. Text messages weren’t considered a consumer product at first. Facebook, even, in the early days, was something entirely else,” Saramaki says. “But people just find a way to use those technologies in ways that are meaningful to them. And that’s probably related to intrinsic ways of how we want to structure our networks and carry out our social interactions.”
And yet, the PNAS study does contain one glaring loophole. The authors didn’t ask the subjects about who were their romantic partners, among their close relationships. The presence or destruction of a long-term romantic relationship, Saramaki notes, could cause larger fluctuations in social signatures that the study didn’t take into account.