Friendship is likely to be an important determinant of our health, well-being and happiness. Nonetheless, our social world is often small, limited by our psychological ability to manage friendships and our free time. Does online networking allow us to increase the size of our social world?

Professor Dunbar is Department of Experimental PsychologyUndergraduates socialising in the University Parks 

By Professor Robin Dunbar

Social media from Facebook to WhatsApp have revolutionised our social lives in less than a decade. Yet, claims to the contrary notwithstanding, they haven’t in fact increased the size of our social circles. These remain resolutely fixed at a typical value of 150 (with a range from about 100-250, reflecting the difference between introverts and extraverts). Why is this? The main reason stems from what has become known as the Social Brain Hypothesis, the explanation for why monkeys and apes have much larger brains than all other animals. The size of core parts of the neocortex sets limits to the size of social group you can have. Not only does this limit human group sizes to around 150, but brain scanning studies have revealed that there is a correlation between the size of our individual friendship circles and the size of core regions of our brains, especially in the prefrontal cortex (the region just above the eyes). Department of Experimental PsychologyMany of the social tactics we use to increase the numbers of people we can bond with cannot easily be replicated online

Not only are our personal social networks about 150 in size (which includes your extended family as well as friends), but this figure turns up in many different forms of human social organisation. The average village size recorded in the Domesday book was about 150 in county after county, and seven centuries later English village sizes (as determined from church baptismal and burial registers) were still the same size. It is the typical size of parishes among both the Hutterites and the Amish Anabaptist sects in the USA, the ideal size of Goretex’s factory units, the typical size of research subdisciplines in both the sciences and the humanities, the average size of hunter-gatherer communities, and the average number of friends that people have on Facebook. It is also the average size of the foundational unit, the company, in all modern armies.Students on a bench in the city centre checking a mobile phone,Students on a bench in the city centre checking a mobile phone

However, neither our social world nor most of these very different kinds of human organisation are homogenous structures. If we look at the patterns of interaction that occur within these organisations, we find a very distinctive pattern of structure. They typically consist of layers that have a very specific scaling ratio – each layer is three times the size of the layer inside it. In terms of our friendship circles, we might think of these as constituting intimate friends, best friends, good friends, and just plain friends. And in respect our of personal social networks we know that this layering continues out beyond the 150 to include layers that we can think of as acquaintances (people we work with or know socially that we might occasionally have a drink with, but would never invite home) and all the faces we can put names to. These layers have very distinctive sizes: 5, 15, 50, 150, 500 and 1500, counting inclusively (so the 15 includes the innermost 5 and so on). We see these layers in our face to face interactions, our ratings of emotional closeness to others, in how often we phone people, in our postings on Facebook and even in the frequencies with which twitterati (the followers of Twitter accounts) ‘converse’ with each other.Having a drink at the oldest pub in Oxford, the Turf Tavern, outside in the sunshine Socialising the old fashioned was: having a drink at the Turf Tavern in Oxford

This layering pattern is created by the frequencies with which we contact the various members of our social circles. Because time is limited, we allocate our social effort in ways that maximise the benefits. So we devote most time and effort to the people who provide us with the most social and emotional benefits – the shoulders to cry on. These are the innermost circle in our social network, and typically number just five people, and we devote 40% of our social effort to them. The next layer out (our best friends) includes another ten people (giving 15 people in the two combined layers), and we devote another 20% of our effort to these ten individuals. So 60% of our social time is concentrated on just 15 people who matter most to us. The remaining 40% of our time is spread increasingly thinly among the remaining 135 people in our social networks, with just a little set aside for our acquaintances.

Each of these layers corresponds to very specific frequencies of interaction. Friendships are so dependent on interaction, whether online or face-to-face, that failure to match these rates of interaction will result in friendships deteriorating very quickly and dropping down through the layers to eventually become ‘just an acquaintance”. However, the real problem is that many of the social tactics we use to increase the numbers of people we can bond with cannot easily be replicated online. These include laughter, singing, dancing and eating and drinking together, as well as going to the theatre and films with friends. Social networking sites notwithstanding, these form the meat and drink of everyday relationship maintenance.

Professor Dunbar is a research professor at the Department of Experimental Psychology. He gave a talk on this subject at the 2016 Alumni Weekend.

Images: Pintrest, Shutterstock, Oxford University Images


By Dr. Evangelia Sembou

The truth is that social media are important nowadays. They help people keep in contact. That said, they cannot replace social and/or interpersonal relationships. One may have a huge circle of friends on facebook, but still be the loneliest person in the world. Interpersonal relations can only be face-to-face, while one's social circle is the circle with whom one socializes.

By Laurie Allen

There seems to be some confusion here about the optimal size of our personal social networks. If the 'normal' figure of 150 includes our extended family and friends, then the Goretex and research sub-discipline examples are actually counter-examples - when one takes into account the employees' families and other friends they suggest that people in these organisations have a PSN which is significantly greater than 150.

Or have I missed something?

By Marek Czarnecki

Very interesting to see a purely social (and essentially qualitative) characteristic boiled down to a simple quantitative metric.

Hari Seldon would be proud :-)

But the study seems rather backward looking in that the value of 'cyber' relationships is discounted entirely. Is there not some way in which a remote relationship can move into the inner circles of the model you describe.

Or is it there actually an inherited pressure exerted to build one's personal model, forcing people to ascribe a closer group than is merited to these remote associations.

Or simply, is the penetration of Skype directly proportional to the average misery of the populace ?