Book Notes: Robin Dunbar — Friends

From kindergarten to the nursing home, friends are some of the most important relationships in our lives. This becomes painfully apparent in our current times when many of us had either too little or too much contact with our friends. This book opened my eyes to the many dimensions of friendships and the fundamental role friends play in shaping who we are as individuals and as a species.

Dunbar, R (2021): Friends — Understanding the Power of Our Most Important Relationships. London, UK: Little, Brown Book Group


In this book, eminent Professor Robin Dunbar discusses friendships. His treatise covers the evolutionary origin of friendship, the brain mechanisms of individual differences in social skills, differences in friendship style between men and women, changing friendships across the lifespan, differences in friendship between the online and real-world, among many other topics. The thorough look into friendship is both illuminating and refreshing. There are many books on the psychology of other relationships, e.g. between romantic partners, siblings, or parents and children, but friendships are a neglected topic. The great importance of friendship should be obvious to everyone now that we had to spend a long time either relatively isolated from our friends or cooped up with a few of them. The book uncovers many of the behaviours that highlight why we may be friends with certain people, how we maintain friendships, and why some friendships break down. I found these discussions intellectually stimulating and practically important. The book is also very readable. There are many anecdotes and funny observations sprinkled in that make the more technical discussions more approachable. My only slight criticism is that there are some extrapolations, especially around gender differences and evolutionary psychology, that veer a bit too far from the supporting evidence for my taste. In sum, I think this book is great for anyone interested in psychology, human evolution, or behavioural economics.

Please note: This is my summary of the content of the book. While I tried to capture the main points, this is necessarily a subjective and incomplete overview. I highly recommend that you pick up the book yourself to get the full details.

Friends are important for mental and physical health. People with fewer friends experience more psychological distress. Further, people who are more socially integrated show better markers of physiological health (lower blood pressure, lower body mass index, lower inflammation). These associations are found in children, adolescents, adults, and the elderly.

In a longitudinal study of 267 males, Jenny Cundiff and Karen Matthews found that the more socially integrated a child was at age six years, the lower their blood pressure and body mass index two decades later in their early thirties. This result held up when they controlled for race, body mass index in childhood, parental socio-economic status, childhood health and extraversion.

Dunbar defines a friend as “the sort of people you would like to spend time with if you had the chance, and would be willing to make the effort to do so.” Notably, family members are included in this definition.

We do … typically have around five intimate friends, but we also have 150 friends-in-general.

This number of around 150 friends is known as Dunbar’s number. Indeed, multiple lines of evidence suggest that human social networks fall within the 100–200 range.

The number of friends changes across the lifespan with increases until the early thirties, followed by a stable number of friends and a subsequent decline from 60 onwards. There are also individual differences based on personality. Namely, extroverts tend to have a larger network. However, introverts spend more time with individual friends in their social network.

There is a linear association between the size of the neocortex and the typical size of social groups in apes. This association predicts a groups size of 148 for humans. In other mammalian species, the association is more like a step function. Monogamous species have a much larger brain than solitary or herd animals. The social brain hypothesis suggests humans began to live in groups to deal with ecological challenges. Then, brain size needed to increase to cope with the additional challenges of living in social groups.

Within humans, the volume of the orbitofrontal cortex, dorsomedial prefrontal cortex, and lingual gyrus, and total white matter volume is associated with larger social networks. In turn, loneliness is associated with less grey matter in the superior temporal sulcus. There are some sex differences in the association between neuroanatomy and social network size. Women who live in households with more people were found to have a larger amygdala. In contrast, in men living in larger households, the orbitofrontal cortex was larger.

There are four friendship circles that scale at a ratio of about 3:

  • Close friends (~5): These are emotionally close friends that we spent most of our time with (40% of time devoted to them) and that we contact at least once a week. These friends provide emotional, physical, and financial help and advice when needed.
  • Best friends (~15): These are our everyday social companions, i.e. the people you would invite for a quiet dinner. We typically dedicate around 20% of our time to these friends and contact them around once per month.
  • Good friends (~50): These are our “party friends”, i.e. people you would invite for a BBQ or a birthday party. We typically contact these friends once every 6 months.
  • Just friends (~150): These are the people we would invite for a wedding or who would attend our funeral. These are people we contact around once a year.

The innermost circle can be further subdivided to include “intimate friends” of 1.5 people. For men, that is typically the romantic partner. For women, it is the romantic partner and their closest friend (BFF).

… we devote about 40 per cent of our total social time to the five people in the innermost layer, and a further 20 per cent to the 10 people that make up the rest of the next layer, the 15-layer. That is, 60 per cent of our total social effort is devoted to just 15 people.

Friendships require active maintenance. Part of it has to do with how easily we can engage with friends. It is much more likely that friendships are maintained when we can reach the friend within 30 minutes. If we cannot see friends regularly, the friendship degrades at around one standard deviation for each year apart. For girls or women, talking together is the most important activity to preserve the friendship. For boys and men, doing things together is more important.

There are some differences in friendship patterns between people who are more active in the morning (larks) versus people who are more active in the evening (owls). Owls tend to have a larger social network and prefer to socialise with other owls. In contrast, larks have smaller networks but spend more time with individual friends.

Cognitive abilities were important for the expansion of the social network in humans. Further, individual differences in cognitive ability are related to differences in friendship patterns. One important aspect is the ability to understand other people’s perspective, i.e. Theory of Mind. Theory of Mind is important to interpret other people’s intention, resolve ambiguities in their language use, and lie to them convincingly. The better a person’s Theory of Mind, the more friends they tend to have.

Joanne Powell was able to show, using a statistical technique known as path analysis, that the causal chain has to be: brain size determines mentalising ability, and mentalising ability determines the number of friends you have.

Inhibition is another important cognitive ability for friendships. People with better inhibition tend to also have better social skills, which, in turn, is related to the number of friends they have.

Touch is an important medium for social bonding in humans. Our tolerance for being touched indicates how close we are to another person. Being touched stimulates specialised fibres in the skin (afferent c-tactile neurons) that, in turn, activate the endorphin system. The endorphin system is implicated in attachment style, empathy, and to some degree in romantic relationships. Other neuromodulatory systems also play important roles in social behaviour. The oxytocin system is particularly important for pair bonding. In contrast, the dopamine system is related to the number of close friends a person has and how to engage they are in their community.

The closeness of friendship depends on the time people spend together. It takes around 45h to move from acquaintances to causal friends, an additional 50h to move to meaningful friends, and another 100h to move to best friends.

One-on-one social grooming was probably too time-consuming to maintain the many social connections that are necessary for human social networks. It seems that humans developed additional methods to bond with groups of people. One such method is laughing together. Laughing triggers the endorphin system, makes people more relaxed, and creates a sense of bonding with the people one is laughing with. Dancing in synchrony, singing together, and eating and drinking together have a similar effect, i.e. these activities activate the endorphin system and create a sense of bonding.

Social eating tends to influence the number of close friends we have and our sense of life satisfaction most directly, …, whereas drinking socially influences our sense of engagement with the community and our trust in its members most directly, …

Language is likely to have played an important role in the expansion of the social network in humans. Language fulfils at least two important functions. First, we can track the behaviour of more people through language than through direct observation (gossip theory). Second, language can be used to create narratives.

Stories and folktales are what binds us together as a community

How important social aspects are is evident from the observation that people typically spent ⅔ of their conversations on social topics and tend to remember social aspects better than factual information. Further, time spent in meaningful conversation with others is a strong predictor of life satisfaction.

People tend to befriend people who are similar to them (homophily). Friends tend to share more genes with each other than with random strangers and show more similar brain activation patterns when watching a film. The similarity of friends can be broken down into seven pillars:

  • having the same language (or dialect)
  • growing up in the same location
  • having had the same educational and career experiences
  • having the same hobbies and interests
  • having the same worldview (moral, religious, and political views)
  • having the same sense of humour
  • having the same musical tastes

The more of these pillars friends have in common, the closer they tend to be. We typically share 6–7 pillars with friends in the innermost circle, but only 1–2 pillars with friends in the outer circle. In addition to the pillars, we also tend to be friends with people of a similar age, gender, ethnicity, and personality.

..our choice of friends is heavily dictated by trying to find like-minded people, people we feel comfortable with in casual company, people we don’t have to explain the joke to every time, people who think like us and whose behaviour we don’t have to work hard at trying to understand, people with whom conversations have a natural and effortless flow that we don’t have to work at.

… relationships are like a game of snakes and ladders. Trust builds up over time as we have more and more positive experiences with someone. Then they do something that upsets us, and the relationship slides down a snake to land a row or two lower on the trust stakes.

The most important relationship changes across the lifespan. For females, other females are the most important contact for those under 18 and over 40, but from 18 to 40, it is likely a male. For males, the pattern is similar, but the opposite gender orientation only lasts until about 30. This shows that women have a longer focus on their romantic partner than men.

Satisfaction with the romantic relationship was associated with having a similar personality and similarly high levels of self-control.

Around 70% of men and women prefer people of the same sex as friends. That is because they feel that same-sex friends provide more help and greater loyalty. Men and women tend to behave differently in friendship. Women treat friends like a sibling and enjoy time spent in conversation with them. When experiencing stress, women engage in behaviours that promote sympathy and approach from their friends. In conversation, women are cooperative and use active listening comments. In social conflict situations, women engage in psychological fighting aimed at undermining the reputation of the component and tend to hold grudges when they are wronged.

Men treat friends more like cousins and enjoy joint physical activity. Men respond with distancing behaviours when experiencing stress. The conversations of men are competitive with frequent one-upmanship. In conflict situations, men fight physically when they do fight. Men have a higher tolerance for stresses that occur in relationships and are more ready to forgive when they have been wronged.

For women, best friendships were most strongly predicted by the degree of similarity in education, sense of humour, dependability, a happy disposition, number of shared activities, how much mutual support was exchanged and, specifically for same-sex best friends, the frequency of contact via digital means (phone, Facebook, email, etc.). Interestingly, a shared history had a negative effect on intimacy in these relationships. For men, best friendships were best predicted by the duration of the friendship, having a shared history, the amount of mutual support exchanged, the number of shared activities and the degree of similarity in financial status (presumably important for sharing rounds at the pub or other social events), outgoingness, dependability and number of social connections. These results held irrespective of the sex of the best friend.

Friendships die when we do not see the people concerned often enough to maintain the relationship at its former level of emotional intimacy — and especially so when neither side can quite muster the energy to do anything about it. So the tendency is for such relationships to fade quietly, almost by accident rather than design.

Friendships frequently end. There is 10% turnaround of friends in the inner circle per year. Most friendships break down within the first 3 years. Relationships with family are the most protected (kinship premium) and female-female friendships are the most fragile. The higher breakdown rate of female-female friendships is likely due to the higher expectation of women. There are some essential rules for maintaining friendships:

  • standing up for a friend in their absence
  • sharing important news
  • providing emotional support when needed
  • trusting and confiding in each other
  • volunteering to help when needed
  • making an effort to make the other person happy

If the rules are repeatedly violated, the friendship may break down. If reconciliation happens, it typically happens within the first week. The most successful strategies for reconciliation are apologising, having a frank discussion, paying financial reparations, or taking a relationship break.

Friendships change across the lifespan. Children only begin to form friendships at around 5 years of age. Before then, they apply the label “friend” irrespective of reciprocity, do not feel social exclusion to the same extent as older children, play more in parallel than together with other children, and are not aware of their position within the social network. In later childhood, children begin to form friendships. Boys tend to prefer playing in larger groups, while girls prefer one-on-one interactions. Girls report more relationship breakdowns with friends and suffer more from them than boys do.

In old age, the number of friends in the friendship networks tends to diminish. This is mostly due to a decreasing number of social friends. The family component of the network remains relatively stable as we age. The structure of the network changes, too. People tend to sacrifice casual acquaintances as they age and focus more time on closer friends.

… there seems to be a divergence in network size with age between the two sexes. Women became more social, men less so.

Despite the seemingly endless possibilities to connect in the online world, people show many of the same characteristics that are also apparent in their real-world social interaction. For instance, people tend to have a similar social network size online and in the real world, and communicate mostly with people who they also know in real life. Further, the same gender-typical behaviour is apparent online, e.g. women tended to form alliances with other women in a multiplayer online game.

There are some differences in social behaviour depending on the communication channel. Messaging services are mostly used to communicate with close friends, while social media is used to stay in contact with friends who live further away. An association between medium and social aspects is also apparent in more traditional media. Text messages are mostly used for good news and phone calls for bad news. We also tend to call family and text with friends.

There are some potentially negative aspects to online communication as well. In a group of adults who moved between cities, the ones who spent more time online made fewer friends in their new location and showed an increased incidence of loneliness and depression. Further, more passive social media use is associated with higher anxiety and depression, but active social media use had the opposite effect.

If children spend a lot of their time online, they may not be gaining the experience they should be having in two important respects. First, most of their interactions online are dyadic rather than involving many individuals in a group. Second, if someone does kick metaphorical sand in their face, they can simply pull the plug; there is no obligation to learn how to compromise.

I’m a lecturer in psychology specialised in cognitive neuroscience. My research investigated brain development in young people who struggle in school.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store