Falling in love feels like your life is getting bigger, but Robin Dunbar's research suggests it usually costs you about two close friends, because the inner circle most people can sustain holds only around five, and a new partner has to take someone's place

Five. If you've ever sat through a social psychology class, you've probably heard that number more than once. I heard it so often I could probably recite it half asleep. According to Robin Dunbar, most of us can keep only about five people truly close: the ones we'd turn to when everything falls apart. Then I read a paper by Dunbar himself saying that falling in love costs you nearly two of them. Time to test the theory I'd been taught on repeat, I thought.

Spoiler: the five held up. The "two close friends" part, less so.

So where did the two friends go?

The paper is by Max Burton-Chellew and Robin Dunbar, both at Oxford at the time, and it came out in 2015 in Evolutionary Behavioral Sciences. Its title doesn't hide the verdict: "Romance and Reproduction Are Socially Costly". The idea had been around for a while by then: Oxford's student paper, Cherwell, was already reporting it in 2010.

Here's what they actually did. They asked 540 people online to list everyone they "could approach for help in times of 'severe emotional or financial crisis'", up to 15 names. On average, people named 5.25. So yes, the famous five again.

Robin Dunbar on why friendship matters, in a 10-minute talk for the British Academy (April 2021). From 6:18 he describes the inner core of about five, his "shoulders to cry on" friends.

Then they compared people who were single with people in a relationship. Among those under 36, singles named 6.2 people on average. Those in a relationship named 5.4. That's a drop of 0.8, which doesn't sound dramatic. But 70 percent of the people in a relationship had put their partner on the list. Take the partner out, and the others had lost about 1.6 people, with the partner filling roughly one seat. In the paper's own words, "the cost of romance is the loss of nearly 2 members."

Notice what the question was, though. It asked who you would call if your life fell apart. That list overlaps with your close friends, but it's a different list.

Plot twist: they were mostly family

This is the detail I didn't see coming. The people who dropped off weren't mostly friends. Among the under-36s, relatives on the list went down by 1.0, from 3.6 to 2.6. Unrelated people, partners not counted, went down by 0.6, from 2.6 to 2.0. The authors' reading is that a new partner "typically displaces a close relative from the support network as well as an unrelated friend."

In other words, the researchers' own numbers point to about one family member and half a friend. Less catchy, I admit.

A few more things the headline left out:

• It only held for people under 36. For older people, the pattern flipped, because single people's circles seemed to shrink with age.
• Having children shrank the circle further, more than the relationship itself did.
• 32 people, about 9 percent of those in a relationship, listed a second romantic partner as someone they could count on. Only five of them also listed their main partner.

That last one isn't about friendship at all, but I couldn't leave it out.

Aristotle got there 2,300 years earlier

Long before anyone counted names on a list, Aristotle was making the same point in the Nicomachean Ethics. He thought the best kind of friendship, the kind built on who the other person is, takes time and closeness, so you can't have it with many people at once. And he compared it to love:

"One cannot be a friend to many people in the sense of having friendship of the perfect type with them, just as one cannot be in love with many people at once (for love is a sort of excess of feeling, and it is the nature of such only to be felt towards one person)."
Aristotle, Nicomachean Ethics, Book VIII (translated by W. D. Ross)

Diane Felmlee, a distinguished professor of sociology and demography at Penn State whose research looks at friendship and dating, put a modern version of this to me: "It takes time and attention to cement a romantic relationship, especially a new one. It is difficult to balance fully engaging in a romance, and at the same time not only maintain old friendships but take the time to develop a new friendship."

Aristotle also described lighter friendships, the ones built on being useful to each other, and said those end easily once the use runs out. If you've ever lost touch with work friends a month after changing jobs, you know the type.

And here's the other side of it: Aristotle thought the friendship between husband and wife "may be based also on virtue, if the parties are good": a partner can be a friend of the best kind too. Which is exactly the question the 2015 paper leaves open. Does one very supportive partner make up for the people who drifted off the list?

When I fell in love, everyone else went quiet

I can answer that one from my own life, at least for me. When I first got into a relationship, I didn't lose just two people.

Everyone else went quiet.

I didn't feel the need to share anything with anyone.

Then things weren't right anymore. And it was really hard to find someone to talk to. Not because people weren't there for me. They were. I just wasn't used to sharing with anyone else anymore.

So did one partner make up for the smaller circle? For a while, yes. Until the day I had to accept the fact that I maybe needed that circle.

Do the friends ever come back?

I've been on the other side too. When my best friend got into a relationship, she stopped sharing everyday things with me. She didn't notice. I did, and it felt empty. A few years later, though, she started telling me the small daily stuff again. I'm still not sure why. Maybe the honeymoon phase in her relationship was simply over.

Is that typical?

A much bigger study can help here, because it followed the same people over the years. Matthijs Kalmijn, a professor at the University of Groningen, used the Swiss Household Panel, which interviewed thousands of people every year for 12 years, to see what happened when they moved in with a partner or married. His finding: "the number of friends does not decline after union formation", but "the amount of contact with friends declines considerably (by 23% for men and 28% for women)." To me, his explanation sounds a lot like Aristotle: spouses and good friends "fulfill similar functions".

Teenagers show something similar. In a study of Glasgow teens, Haoyang Zhang and Felmlee found that those who were dating were "not more likely than those who were unattached to dissolve their friendships". They were, however, "significantly less prone to form new friendships over time." As Felmlee put it to me: "I believe it is much easier to fail to make new friends than it is to end existing friendships."

So maybe love doesn't push friends out so much as turn the volume down. They stay in your phone. You just call them less, and the new ones never quite make it in.

A word of caution before anyone blames their partner. The 2015 paper compared different people at one moment; it didn't follow anyone as they fell in love, and the authors say so themselves: "Our study was not longitudinal." Less sociable people might simply be more likely to be in a relationship in the first place. The 540 people were volunteers for online psychology studies, most of them women and fairly young, which is common in research but not the whole population.

Still, the five keep coming back to me, the way they did in every lecture. So here's my question for you: who's in your five right now? And who quietly stepped out when someone new stepped in?