Skip to main content
Menu
How Many People Can You Remember?

Dear Joe,

First up, let me try to look at faces.

It’s generally accepted that people are better at remembering faces than names because a person’s mug is so rich with visual information (how many times have you seen someone and struggled to remember her name? how many times have you remembered someone’s name but struggled to remember what she looks like?). But studies haven’t tried to determine the exact number of faces people can remember. And to be fair, I really can’t imagine how any scientist could design an experiment that would do so reliably.

Studies have looked at what makes some faces easier to remember than others. A 1999 paper published in the Psychonomic Bulletin & Review found that familiar faces (meaning faces that people had seen before) were easier to remember. That doesn’t sound too surprising, although considering the images that were used for the experiment — shown below — it’s kind of incredible that participants remembered any faces at all.

dearmonafaces

In another study, published in 2003, researchers at Colorado State University analyzed human face recognition and found that certain features had a statistically significant effect on participants’ ability to recall faces. Faces with closed eyes, bangs or facial hair were easier to recognize — as were older faces. And the researchers noted higher recognition rates for male faces (women were literally easier to overlook — pretty depressing, eh?).

When it comes to names, things get even more complicated. Again, try to put yourself in the position of a social scientist for a minute. How would you design an experiment that looked at the total number of names an individual could recall? I guess you could ask her to list the names of every person she could remember — but then you might not capture all the Bridgets and Bernies whose names the participant only remembers once she bumps into them on the street. And in a way, that’s kind of closer to the way real life works, right?

Most research has focused instead on how many people you know and can remember (along with their names). I’ll share some of it here because it’s pretty interesting and still relevant to your question.

Ithiel de Sola Pool and Manfred Kochen were two sociologists who, in the 1950s, pioneered research on “acquaintanceship volume” — estimating the number of individuals people have in their social networks. To measure his own acquaintanceship volume, Pool carried a notebook around with him for 100 days and whenever he exchanged words (in person, over the phone or by mail) with someone he had previously met and whose name he knew, he noted the person down in the book. Each person was noted only once even if Pool interacted with him or her multiple times, so as the days passed, the growth in new names slowed. Pool then used the data to predict how many names his notebook would contain if he conducted the experiment for 20 years. Pool estimated that he would have recorded about 3,500 acquaintances by the end of two decades.

That might sound like a crazy-high number, but in 1960, an MIT student looked at 86 days of President Franklin Roosevelt’s appointment book and estimated that Roosevelt probably had about 22,500 acquaintances. In 1961, Michael Gurevich repeated Pool’s diary experiment with 27 people and found that the average number of acquaintances predicted after 20 years was 2,130.

I bet I know what you’re going to say though, Joe: Writing down names isn’t the same thing as remembering them, right? Well, Pool thought of that too. He wanted to test whether he could remember all his acquaintances. He did this by using two phone books (one for Chicago and one for Manhattan) as prompts. After randomly selecting 30 pages from each, he looked over the names on those pages and tried to think of people he knew with the same family name. He ended up remembering 3,100 acquaintances with the aid of the Chicago book and 4,250 names with the Manhattan phone book.

No one would look at a phone book now. And our closest equivalent, social media, could be influencing our memory. If anything, Facebook, Instagram and Twitter prompt you to remember more people than you would otherwise because you see their names and faces more often in your feeds.

Maybe you’re not looking to remember people as much as you’re keen to remember their names, Joe. If so, I can offer you some advice, courtesy of Richard Harris, professor of psychology at Kansas State University. You can repeat the person’s name back to her while you’re talking to her, Harris says, “although the best strategy is simply to show more interest in the people you meet.”

But maybe there are some people you’d rather forget — I have Benedict Cumberbatch’s face and name etched into my head whether I like it or not. In that case, it might make more sense to look not at how many faces (with their corresponding names) you can remember, but instead at how many you might want to remember — because they’re your friends, the type you would invite to your wedding. The answer, according to Robin Dunbar, is 150.

Dunbar is a professor of evolutionary psychology who, in the early 1990s, was studying primates and the size of their social groups. He wanted to know how many individual relationships the primates could maintain within their larger social context. Using a formula based on brain size, Dunbar estimated that the Macaca sinica (pictured below) tends to run in groups of around 17, while the Cacajao tends to have about four monkey pals. When Dunbar applied his formula to humans, he predicted that the typical social group size — that’s the largest number of individuals that we humans can maintain stable relationships with — would be 147.8 (to be precise, he estimated it would be somewhere between 100 and 231 people).

Macaca_sinica-Sigiriya_Village-Sri_Lanka_(3)

By Ji-Elle

To test his theory, Dunbar started off by looking at modern hunter-gatherer societies, where he found three levels of social cohesion. On one end was the small living groups or overnight camps (which had between 30 and 50 people); on the other was the large population unit or tribe (which had between 500 and 2,500 individuals). Between those two levels was the clan, which typically contained between 100 and 200 people. That’s darn close to Dunbar’s estimate of the number of relationships people can keep track of. (Clans, according to Dunbar, “interact on a sufficiently regular basis to have strong bonds based on direct personal knowledge.” You might feel like you have “direct personal knowledge” of Benedict Cumberbatch, but Dunbar would probably disagree — his number doesn’t include celebrities.)

It’s not just clans, though. When Dunbar looked at the smallest independent unit in various armies, he found that the average size was 179.6 — again, within the bounds of his original estimate of social group size. And in a follow-up study he wrote in 2002, Dunbar found that Christmas cards were on average sent out to 153.5 individuals.1

From all this, Dunbar inferred that “there is a cognitive limit to the number of individuals with whom any one person can maintain stable relationships.” So, chances are, there are about 150 people whose names and faces you can remember without a prompt — and a hell of a lot more acquaintances that would come to mind with the right encouragement.

Hope the numbers help,

Mona

Have a question you would like answered here? Send it to @MonaChalabi or dearmona@fivethirtyeight.com.

More from Dear Mona:

When Will Everyone I Know Be Married?

What Does It Mean To Die Of Natural Causes?

Does It Make Sense To Split The Check At A Restaurant?

Footnotes

  1. The average number of Christmas cards mailed was 68.2 — 153.5 was the number of individuals living in the households where they were sent.

Mona Chalabi is data editor at the Guardian US, and a columnist at New York Magazine. She was previously a lead news writer for FiveThirtyEight.

Comments