In the previous chapter I developed the idea that the modern world is full of stresses and that many of these stresses have come about recently. A good proportion of the stresses we met during our ‘typical’ modern day were focused on something that virtually all of us now take so much for granted that it becomes very difficult to imagine life without it: the mobile phone. Whether it is ‘hassle’ from social network notifications or the ‘always on’ culture that phone-based email and enhanced connectivity promotes, it is clear that the mobile, and the virtual world that it enables us to connect with, might be problematic. It is not surprising that we are having trouble coping with this revolution because our phones connect us to the world in ways that even our most recent ancestors (and unless you are a very young reader, I am talking about our parents) could not imagine. Before considering the problems that such technology can cause, and why they stem from an evolutionary history that has adapted us for a very different world, it is worth taking a moment just to realise how powerful those beeping, vibrating, stressful, distracting hunks of plastic, metal, silicon and rare earth minerals really are.
If you are aware of any computing in the 1980s you will likely have heard of the Cray-2 supercomputer. In 1985 it was the fastest machine in the world, an absolute beast of a processor used for the sort of applications and organisations that get thriller writers all of a shiver: nuclear weapons research and secret sonar testing for NASA and the United States Government. It was used for other things as well, like designing Ford cars and for less classified research in universities. The Cray-2 was the powerful computer of the mid-1980s and yet the iPhone nestling in your pocket is far more powerful. Even the now outdated iPhone 4 had nearly three times the processing power of the mighty Cray-2, and there is absolutely no doubt which device had a better video camera or produced better graphics. Also, with its hefty footprint and need for a complex built-in cooling system the Cray-2 was not exactly pocket-friendly. You might complain about the power consumption of your mobile phone but the Cray-2 needed 200kW, equivalent in car terms to 268 brake horsepower or about the power of a full-spec Land Rover Discovery Sport. In 1969 NASA put Neil Armstrong and Buzz Aldrin on the Moon and guided Michael Collins (in my opinion the most interesting of the three Apollo 11 astronauts) around the Moon to pick them up again using a combined computing power that could be matched by just two Nintendo Entertainment Systems, used by the serious gamer back in the Cray-2’s heyday. The latest mobile phones have processing power that is so much greater than NASA in 1969 that it’s hardly worth comparing, but if you’re interested, the nearly 30kg (66lb) guidance computer onboard the Command Module had a processing speed of 0.044MHz; the iPhone 5 runs at 1.3GHz or about 30,000 times faster. The code required to monitor the status of the flight and the astronauts took up 6MB; the iPhone 5 can store 1GB or about 167 times more, and yet you use it to ‘like’ a video of a cat sneezing. It’s almost insulting…
Powerful processors and massive memory are undeniably impressive but ultimately they are only measures of the potential of phones. It is the penetration of mobile phones into our everyday lives, their ubiquity and, frankly, their absolute necessity that are the true measures of their success. Virtually all of us have them, they are everywhere and they can do pretty much everything. The original function of a telephone was to convey messages using speech, which was a step change in technology from the telegraph, but telephones quickly changed from being conveyors of ‘important messages’ to instruments for casual conversation. Likewise, mobiles have gone from workaday objects capable of connecting us on the move, the ultimate 1980s accessory for any yuppie with enough arm strength to hoist it to their ear, to the defining technology of our time. They have also become the primary tool that we now use to construct something central to this chapter and central to our modern lives: our virtual social networks.
The rise of social media
Unless you’ve been living truly off grid for the last 15 years or so, you cannot have escaped the term ‘social media’. For a start, many news stories via ‘traditional media’ these days come adorned with the thoughts and opinions of a coterie of celebrity Tweeters. Quite often these celebrities are virtually unknown outside of the virtual world in which they are lauded, having become elevated to the status of ‘influencer’ because of some beautiful Instagram shots attracting thousands of followers or some tweets that happened to ‘go viral’. Insta-famous, influencers, tweets, Facebook-friends, ‘likes’, going viral; literally none of those words and phrases would make any sense to us even 15 years ago. A primary reason for this great rise in social media is the fact that we can interact with the virtual world simply, reliably and cheaply through our smartphones. If we had to log on to a desk-bound computer every time we wanted to update our status or post a photo, uploaded from a digital camera via a lead or card slot, it is hard to see too many of us bothering. Certainly no one would be going through the effort required multiple times an hour. Yet now, because of our phones, for many people the virtual worlds that we now inhabit are as important (or more important) than the very different ‘actual world’ in which we evolved. Throw in some super-realistic massively multiplayer gaming, the dark web and virtual economics and it’s hardly surprising that we’re rapidly getting out of our depth.
Before I get too far out of my depth, I should define the battleground here. I’m mostly concerned with the aspects of the human animal that have evolved for a world in which we no longer live, or those aspects that were rather useful in our evolutionary past but are coming back to bite us now that we have altered our environment. The science exploring our interaction with the virtual world is still, obviously, very much in its infancy. The psychology of those ‘lost’ in highly immersive virtual worlds is fascinating, and surely linked with our propensity for abstract thought and our addictive tendencies (Chapter 8). However, it is our relationship with more everyday social networking, and particularly the size of those networks, that turns out to be the aspect best explained within the context of our evolutionary heritage. It is also the aspect of the virtual world that affects most people across the widest demographic because there are 2.38 billion Facebook users and 1 billion active Instagram users (people who use the platform every month). Twitter is fast falling out of favour if social media commentators are to be believed, but even Twitter has more than 320 million active users, roughly the equivalent of the entire population of the USA. Social media provides an ideal context to explore our relationship between our evolved selves and the massively connected world that our technology has created. At least it is for now, but as we’ve already seen it is a fast-moving world.
At its simplest, social media describes any interactive computer-based application that allows people to create and join networks of other people with shared interests, ideas or need for information, or for pretty much any other reason that people like to connect with each other. Some of the reasons people want to connect are wonderfully wholesome, with networks of people forming on platforms to appreciate the arts, music and books for example. Facebook, the most used of all the social networking platforms, is used by billions across the world to keep in touch with old friends and to form networks of new friends, many of which may never have physically met. There are, of course, many other less wholesome reasons why people want to connect with others and form virtual social networks. There may be a desire to share extreme and possibly outlawed political ideologies, find sexual partners, browse (is that the right verb?) pornography or trade in illicit material and information (child pornography, illegal wildlife products, bomb-making instructions, drugs or guns for example).
I interact on Facebook with a number of people in a manner that would be described as a friendship and yet I have never met them; we are connected through other friends and through shared interests. Likewise, I ‘know’ a large number of people via Twitter that I have never met in real life. For me, and for many others, these online virtual relationships can be very useful and fulfilling. Professionally, I have developed contacts via social media that have resulted in research collaborations, broadcasting opportunities, speaking engagements and, now that I think about it, even put me in touch with the commissioner of this book. Social media can be, and often is, a very important and valued component of our lives. I can honestly say that Twitter and Facebook have enriched my life and I enjoy using them. However, judging from the number of people I now see posting that they are ‘giving up’ Facebook, or ‘taking a break’ from Twitter, it is clear that not everyone feels the same.
Is social media harmful?
The idea that social media and online social networking might be harmful to us in some way began to form almost as soon as social media started to take hold. In 2006 the social media landscape was very different to the one that is familiar to us now. Facebook wasn’t launched publicly until September of that year, Twitter had only launched in July 2006 and Instagram was still a digital twinkle in the eyes of Kevin Systrom and Mike Krieger, who launched the company four years later. Before the Big Three came to dominate in later years, two of the biggest players were Friendster and Myspace. Friendster was a social networking site launched in 2002 that was rapidly eclipsed by Facebook, morphing into a social gaming site in 2011 before being suspended in 2015 and then finally being put out of its misery in 2018. I’ll be honest, the first I’d heard of it was in reading around for this section. Myspace, although largely unknown to younger users now, was the largest social media platform in the world between 2005 and 2008. Unlike Friendster, I had heard of Myspace and indeed once had a Myspace account. Like Friendster though, it proved to be another victim of the relentless rise of Facebook, although it does still exist, at least at the time of writing. Stylised now as ‘myspace’, it has suffered from a string of setbacks in recent years including a major data breach and the loss of 50 million songs and 12 years of uploaded user content.1
Despite their lack of long-term success, back in 2006 both Myspace and Friendster were helping to define social media as we know it and their early success in the sphere was attracting interest from social scientists keen to explore the wider consequences of online social networking. Researchers from the Netherlands studied adolescents using Friendster and Myspace, and they identified two potentially troubling aspects of online social networking that have continued to be a cause for concern over the decade and a half that followed.2 Specifically, they examined well-being and social self-esteem based on feedback that the adolescents (aged 10 to 19) using the social networking sites received from their online friends. What they found out seems now to be obvious. Positive feedback on their profiles (which these days we would refer to as positive comments, likes, shares, retweets or increased followers depending on the platform used) enhanced users’ well-being and self-esteem while negative feedback (negative comments or a lack of interaction) had the reverse effect. Both positive and negative effects were increased with greater use of social networking sites. Even then, in the early days of social media, users were riding the wave of positive feedback, building up self-esteem and well-being through the online validation heaped on them by their virtual network, but also spiralling down in a ‘crash’ of negative feedback. As a Twitter user, I will admit freely that having posts that stack up likes and retweets, getting favourable comments and attracting the attention of ‘super users’ with millions of followers is a bit of a high. It is certainly addictive and I can well imagine how the reverse would also be true, although luckily I am adept at ignoring or ‘muting’ those who deal in negativity. Others are not so fortunate, as is easy to see on any social platform where people have been caught up in a maelstrom of hate and scorn.
It is interesting to see that early investigations into the potential for social networks to cause harm examined self-esteem and overall well-being. These aspects have not gone away in the last decade and a half, with self-esteem worries exacerbated even further by ever more glamorous status updates on Facebook and the rise of Instagram. Instagram is essentially a photo-sharing platform, but the easy availability of filters that can turn even an average holiday snap into something far more attractive considerably elevates it above other photo-sharing sites like Flickr. When all is said and done, Instagram is really about posting photos that say ‘look at how amazing my life is’ and the ‘Insta-fabulous’ lifestyle is immediately clear to anyone who joins the platform. Intuitively, conscious and subconscious comparison of the seeming tedium of one’s own day-to-day life with the heavily filtered highlights of everyone else’s makes blows to self-esteem seem inevitable, even for the sturdiest of egos. For once, the evidence supports our intuition. Surveys of social media users report that social media sites make more than half of users feel inadequate or unattractive. Meanwhile, studies of selfie viewing (people looking at images of themselves that they have themselves taken) found that it lowered self-esteem. Frequent selfie viewing was actually able to lower users’ life satisfaction, especially for individuals with a high need for popularity. Interestingly, ‘groupie’ viewing, looking at images of groups of which you are a part and are being taken by someone else in the group, had the opposite effect; it increased self-esteem and life satisfaction.3 Selfies encourage direct comparison with others in what is termed ‘upward social comparison’, a tendency to compare yourself with people whom you perceive as being better. Groupies on the other hand don’t encourage this comparison. Instead, presumably viewers focus more on the positive feeling associated with being part of that group (usually of friends) rather than the negative feeling of not being someone else.
Further studies on the negative effects of social networking and social media fleshed out the same story and conventional media, threatened perhaps by what might have appeared to them at the time as an existential crisis, were all too keen to fan the flames. Stress of course featured heavily, along with its bedfellow, anxiety (Chapter 5). Add into the mix depression, addiction (more of which in Chapter 8), sleep deprivation (which of course leads to stress and anxiety), relationship crises (there’s some more stress for you), envy and loneliness, and it would seem most of modern life’s ills have at some point in recent years been blamed on social networking. The evidence supporting some of these associations is, you will not be surprised by now to learn, complex. Not only can it be hard to quantify reliably some of the effects, but also monitoring and standardising social media use and controlling for the effects of underlying conditions mean that the sort of broad-brush conclusions that we like are very hard to reach.
Depression is perhaps the most complex of the purported negative associations with social networking. Certainly, some studies have found links. In 2012, for example, two studies investigated the link between depression and social media use in more than 1,700 subjects with an average age of around 19–20, rather endearingly referred to as ‘youths’ by the researchers. The questions we can ask can be dressed up in all kinds of fancy sociological and psychological clothes but fundamentally what we want to know is, does social media use cause depression? The researchers concluded that, with regard to frequency of social networking, the answer to this question is no; they found no evidence that more frequent usage is associated with depressive symptoms. This is superficially reassuring, but the researchers probed a little deeper and were able to reach a more nuanced and frankly more disturbing conclusion. Negative interactions while social networking were associated with increases in depressive symptoms over time, building up feelings of worthlessness, hopelessness and low mood cumulatively with use. They frame social networking as ‘a salient venue in which youths experience the depressogenic effects of poor quality relationships’; in other words, yet another place to get dumped on by the world.4 When this finding is combined with another from the same research, some tentative evidence that some young people may be more at risk than others, then alarm bells really start ringing. Young people prone to ‘depressive rumination’, focusing on symptoms of distress and their possible consequences leading to an unhealthy fixation on negative feelings and problems, may be more likely to experience depressive symptoms as a result of negative interactions on social media. Such individuals may also tend to feel more depressed following social networking interactions. So, while there may not be a clear-cut relationship between social networking and depression, there are clear signs that for some of us social networking can be a contributing factor in degrading mental health. Perhaps worth thinking about, the next time you decide to make a comment on some stranger’s post?
Another study, in 2016, examined the links between depression and anxiety in 1,700 people who used multiple social media platforms. Their results were very clear: people who used multiple platforms were at a threefold risk of depression and anxiety. This shocking headline figure, however, is actually a result of comparing people who use between seven and eleven platforms, which is frankly an awful lot, compared to those who use between zero and two, a number that might be considered more normal. Some of the reasons suggested to account for this increase in depression include having a distorted view of other people’s lives, feeling like time spent on social media is a waste and cyber-bullying. Cyber-bullying, which is not likely to be going away any time soon, is an area where links between social networking and depression, and the role of depressive rumination,5 have been clearly seen in young people. This should be of great concern given that young people have taken their own lives as a consequence. Young people killing themselves (perhaps the ultimate maladaptive human behaviour) as a consequence of interactions in a virtual world of social networks would, 20 years ago, have been the plot line of a dystopian future novel. Now, it is an understandable and serious concern of many parents as their children navigate the world we have created. It’s worth saying this again: we have created a virtual world where young people are killing themselves (and, to take an evolutionary angle, leaving no offspring) because of interactions in that artificial world.
Aside from depression, there is evidence linking anxiety in general to social media use, although there are also studies suggesting that social media can reduce anxiety.6 Social media and mood seems more straightforward, at least according to one study that concluded that Facebook activity negatively correlates with mood; the more you use it the worse you feel. They did not find the same effect with internet browsing, so it isn’t simply a case that being immersed in the online world or screen-time per se are ‘bad’, rather it is something about the social networks of Facebook that causes low mood. They concluded that the reason for this is that people feel bad because they have wasted their time on Facebook. Interestingly, the negative effects of use don’t stop users making what is described as a ‘forecasting error’, with people expecting to feel better after engaging with Facebook despite experiences clearly to the contrary.
Anxiety, stress and depression can all lead to poor sleep, and poor sleep can cause stress and anxiety. Throw social networks into this toxic nocturnal feedback loop and you have another medical issue: insomnia. A study of 18–30-year-olds found a link between social network use and sleep disturbances, with the conclusion that the blue light emitted by smartphones had a part to play, but how often participants logged on to social media sites, rather than the time they spent on them, was a better predictor of disturbed sleep. The results suggest obsessive checking of social media is a problem, but they couldn’t determine whether social media causes disturbed sleep or whether people with disturbed sleep spend more time on social media.7 Cause-and-effect is a harsh mistress.
It seems fair to say that although the picture is complex, there are some links and relationships between specific disorders and conditions, like depression, anxiety and insomnia, and social network use. But what about more general and non-medical measures of well-being or happiness? Does social networking make us sad? Well, probably overall the answer is no, at least according to one study in the Netherlands published in 2018. The study used data from the Dutch Longitudinal Internet Studies for the Social Sciences (LISS) panel for the years 2012–2013. In this survey, participants reported on numerous aspects of their life including their happiness, social networking site use and a variety of aspects of their lives that enabled researchers to infer the quality and quantity of contact with friends and family. It is this last aspect of the study that threw up something interesting. The study revealed a negative association between the total time spend on social networking sites and happiness for users of those sites who were socially disconnected and lonely. This result held even when the researchers allowed for other factors, like hours spent on other internet sites or household income.8 Once again the pattern seems to be that some people are especially prone to negative experiences associated with online social networking.
The paradox that emerges from the findings of many online social networking studies is that, consistently, research shows that friends make us happier, healthier and lead us to live longer. Our friendships are our real-world social networks and having strong social networks is universally regarded as a ‘good thing’, supported by robust empirical evidence.9 Why then might online social networks, virtual friend groups, be a bad thing? We have already seen some of the reasons put forward to explain this negative association, such as a sense that being on Facebook is a waste of time, but these are not connected with any particular mismatch between the contemporary world and our evolution, at least not in a direct or meaningful way. However, there is an aspect of our evolutionary past that does create a mismatch with the online world, and that might go some way to explain some of the issues we have with the ever-increasing sprawl of our virtual social networks. In a nutshell, our brains are too small to cope. To find out more we need to visit the zoo.
It’s all about grooming
The next time you’re at the zoo take a wander over to the primate section and spend some quality time looking at our nearest biological relatives. I don’t just mean a cursory couple of minutes noting that some of them have weird backsides, although the ‘ischial callosities’ on the buttocks of baboons, macaques and chimpanzees are hard to ignore. I mean actually watch their behaviour. It won’t be long at all until you spot them grooming.
Self-grooming, keeping yourself clean and free of parasites by using the mouth in combination with whatever appendages evolution has provided, is a common behaviour throughout the animal world. As the song goes, birds do it, bees do it and even educated fleas do it. Invertebrates (like bees and fleas) groom to remove dirt, fungal spores, mites and other potentially undesirable hitchhikers from their exoskeleton. Birds and mammals on the other hand have a more complex external surface consisting of skin, which needs clearing of parasites, and further external structures that function best when clean and properly set up. Feathers on birds need to be ‘zippered up’ to provide a better flying surface and fur on mammals needs to be clear of substances that cause matting and a loss of insulation. In addition to removing unwanted passengers like mites and lice, the sometimes complex contortions involved in self-grooming (or preening as it is known in birds) can be a way to distribute oils over the feathers and fur. Birds have a uropygial gland located on the top side of the base of their tail that secretes a substance called preen oil. This oil helps to waterproof and condition feathers, making birds better insulated and able to fly. Mammals, including us, have sebaceous glands that secrete oil that lubricates and waterproofs the skin and hair. Self-grooming, then, is a very common behaviour and once you get your eye in for the tell-tale movements you will start spotting it regularly, whether you are watching spiders or sparrows, butterflies or bears. What is more uncommon in general in animals, but will be very clearly apparent in many of the primates you see at the zoo, is the phenomenon of social grooming: grooming and being groomed by someone else.
Social grooming in primates has benefits that extend well beyond the fact that someone else can reach the parts that you can’t. Certainly it is helpful to have another pair of hands to remove those especially stubborn parasites, clean off that ground-in dirt and untangle those proto-dreadlocks, but social grooming has more subtle effects too. One effect is that grooming can reduce heart rate, calming down both groomed and groomer. This observation (in macaques and likely to be the case in other species too) is surely related to the now well-supported fact that stroking cats can reduce heart rate and blood pressure in humans. Indeed, ‘socially grooming’ your cat has been shown to reduce the chance of stroke, a major killer related in many cases directly to high blood pressure, by up to a third. Similar effects of this value of ‘inter-specific’ grooming, whereby one species grooms another, comes from studies of rhesus monkeys. Familiar humans that groom rhesus monkeys produced the same calming effect of reducing heart rate as was produced when individuals were groomed by another rhesus monkey. It may be the case that the calming benefits of grooming don’t even require another living being in order to be expressed. I once watched a captive colobus monkey apparently grooming a soft toy, devoting attention to the furry rabbit with a look of blissful tranquillity on its face. This was especially notable, since the groomer was a male in a group of males, the majority of which were directing most of their ‘grooming attention’ to their own genitals.
The fact that our closest relatives, the chimpanzee and bonobo, spend 10 per cent or more of their time socially grooming10 and that we have a physiological response to grooming-like activities strongly suggests that we have an important evolutionary history when it comes to social grooming. Watching humans interacting reveals many behaviours that would be considered social grooming in any other primates, from brushing each other’s hair (although as the father of three daughters with long hair I do not consider this to be a very calming interaction) to picking lint off each other’s clothes.
But where does social grooming fit in terms of considering how evolution has left us unprepared for the modern world? One clue to this comes in the broader use of the term ‘social grooming’ within behavioural science. Social grooming is often used to describe not just those behaviours associated with maintaining the condition and quality of skin and fur but also behaviours that facilitate the construction of social relationships.11 Social grooming is all about connecting with individuals in ways that develop and nurture alliances and coalitions, cementing and securing relationships and friendships that provide benefits to those involved in the grooming partnerships. In fact, the consensus across most studies of social grooming in primates is that its most critical function is not in removing parasites or improving fur condition (which may have been the initial selective advantage in its evolution) but in establishing, developing and reinforcing social networks and relationships. To use the sort of language that most primate researchers assiduously avoid, when monkeys are picking at each other, unmatting fur and eating ticks, what they are really doing is making friends. By developing good social networks through the act of grooming, primate groups can become more stable, which could bring benefits to individual members through strength of numbers and competitive advantage. It is better for the individual to be in a better group, and a better group is one that has strong and stable relationships between group members. It also forges alliances within the group, and in a number of species scientists have observed monkeys backing up other monkeys in confrontations in ways that really do seem best described as friendships. These coalitions are closely related to grooming relationships, and grooming can be linked to a great many of the different facets of a group’s dynamics. By understanding the benefits of social grooming from a social angle we can develop just the sort of perspective we need to understand the differences and similarities between picking a hair from a friend’s face and friending them on Facebook.
Social grooming in its wider context includes behaviours that are not connected to the physical act of grooming. Gossip, for example, casual conversation between people about people, is considered to be a form of social grooming and like other forms of grooming it has a beneficial physiological effect. Studies have revealed that gossiping, as opposed to other forms of conversation, reduces cortisol (and so reduces stress – see Chapter 5 for why that is such a good thing) and increases the level of oxytocin. Oxytocin is a hormone that plays an important role in social bonding, sexual reproduction, childbirth and the period after childbirth, when it is associated with maternal bonding and the onset of lactation (we will meet it again when we consider trust in Chapter 9). Gossiping has the twin benefits of reducing stress and bringing people together, acting to bond individuals in just the same way that more obvious physical grooming can achieve.12 Such is the importance of gossip in human behaviour that it has been hypothesised to have played a part in the evolution of language. Robin Dunbar, an anthropologist, evolutionary psychologist and primate behavioural expert, proposed in 1996 that language in humans evolved as a response to increasing group size. The logic behind this hypothesis is that conventional social grooming becomes far too inefficient in larger groups; there are simply too many other backs to scratch and not enough of your back that’s itchy. The benefits to individuals of being in larger groups (being able to defend yourself more effectively, to bring down larger prey and so on) would have made the building of stable alliances important, but the need for ‘manual’ grooming to create and maintain these social networks was simply too time-inefficient to get much else done. The beginnings of language, Dunbar proposed, were a form of cheap and efficient ‘vocal’ grooming, starting with the sort of mutually pleasing grunts that allowed groups to bond more efficiently. Over time, vocal grooming sounds evolved into language, with a myriad of additional advantages shaping its subsequent development.
Dunbar’s gossip-grooming hypothesis has been criticised. First among these criticisms is the fact that manual grooming takes time and effort and therefore provides an honest and reliable signal of friendship that simply cannot be matched by vocal grooming; in other words, talk is cheap. Criticism notwithstanding, the Dunbar language hypothesis has at its heart a feature that is of great interest in unravelling the potential links between our evolutionary heritage and online social networking. Group size, Dunbar argued, was a critical factor in the evolution of language because at some point in our evolutionary history, group size prevented conventional social grooming from playing its crucial role in developing and maintaining real-world social networks.13 Group size is also critical to Dunbar’s most famous work, and the idea that now bears his name: the Dunbar number.
The Dunbar number
The Dunbar number is a measure of our social limitations. It is the maximum number of people with whom an individual can maintain stable social relationships in which individuals within the group know who each person is and how each member of the group relates to the others. It is a measure of our maximum social network size and for humans it works out somewhere between 100 and 250, often stated as 150. Dunbar’s number is basically the number of people you can know and keep social contact with. It isn’t the total number of people you might ‘know of’, and neither does it include people you may have once been close to but no longer associate with. Dunbar described it as ‘the number of people you would not feel embarrassed about joining uninvited for a drink if you happened to bump into them in a bar’. Dunbar’s number is a measure of the number of people in a social network that you can feel ‘comfortable’ with; the sort of network that provides the kinds of benefits that we see from having friends. It’s our social feel-good number.
Dunbar didn’t simply magic this number from thin air. He came to this conclusion by considering us biologically, as a primate, and comparing us with other members of our biological order.14 What Dunbar found was that group size in primates related nicely to the volume of the neocortex of the brain. The neocortex is that part of the mammalian brain involved with ‘higher-order’ functions like thinking, spatial reasoning and language. Dunbar suggested that the reason behind the relationship between group size and neocortex volume across 38 primate genera was that the number of relationships an individual can simultaneously monitor is limited by the information-processing capacity of the brain. This processing capacity is itself limited by the number of neurons (measured using the proxy of volume) of the neocortex, leading us to the obvious conclusion of bigger brain, bigger network. When the maximum network size (which subsequently became known as the Dunbar number) is breached, the group can become unstable and begin to fragment.
By generating a predictive relationship between neocortex size and group size it is possible to work out the group size for humans based on our brain size, and it is this approach that leads us to a human Dunbar number of around 150. Dunbar went on to develop and test this idea and found considerable support for it from a range of sources, including group sizes in Neolithic farming villages, Roman army unit sizes and even career-seeking via social networks.15 It is an idea that is intuitively seductive and seems to tie in with our experiences. If it is true, then it would certainly go a long way to account for some of the issues we might have with online social networking, where we have near-constant interactions with networks of individuals orders of magnitude greater than our evolved social limit. Indeed, the Dunbar number is a widely reported and widely applied measure of our social ability. As the famous American biologist Paul Ehrlich said in an ABC broadcast relating to religion in 2015, ‘we are a small-group animal’ looking to survive in a connected world of billions. When asked to elaborate on this by The Conversation Fact Check1 he added ‘The Dunbar number is ~150, size of hunter-gatherer groups, length of Christmas lists, and so on. My point was we’re a small-group social animal now suddenly (in cultural evolution time) trying to find ways to live in gigantic groups.’ He didn’t feel the need to justify the number further or to provide the usual cautionary qualifiers that litter most scientists’ public utterances. The Dunbar number seems then to be widely accepted; but does it, and its implications for modern life, stand up to scrutiny?
Digging deeper into the Dunbar number
As a descriptive and explanatory tool in the social sciences and humanities, the Dunbar number has been widely supported and influential both in theory and practice. For example, in his excellent book The Tipping Point Canadian journalist and author Malcolm Gladwell describes how the company W.L. Gore and Associates, best known for the Gore-Tex brand of waterproof breathable fabric, runs its business according to the Dunbar number. Its manufacturing is organised in such a way that each product can be accommodated in units employing no more than 150 people, with everyone in that unit being housed together.16 Sizing and organising according to the Dunbar number creates an environment where everyone knows everyone else and their relationships to others in the group, and is available to each other. If the group size exceeds 150 then divisions are initiated to create new ‘Dunbar optimised’ groups. For this company at least the approach has been very successful. In other fields though, the entire foundations on which the Dunbar number is based have come under attack.
The Dunbar number rests entirely on the correlation between neocortex size and group size in other primates. One criticism of this approach is that a great many factors come into play when considering the evolution of the brain and especially the neocortex. Humans are pretty big, and large animals tend to have a large neocortex; a sperm whale neocortex is nearly 10 per cent larger than ours in proportion to its body size. Other factors like territory size, diet and activity patterns through the day have also been found to explain neocortex size just as successfully as group size, and overall the correlations and patterns are complex. A critical review of comparative studies of brain size found ‘a number of substantial problems’ with assumptions, data collection and subsequent analysis that were ‘particularly apparent in those analyses in which attempts are made to correlate complex behaviour with parts of the brain that carry out multiple functions’, which is the neocortex in a nutshell.17 Overall, the general mood seems to be that while social behaviour and neocortex volume may well show a correlation, correlations do not imply causality and this is very often the case when complex, interrelated and highly entangled traits are being considered.
Further cracks seem to appear when the anthropological evidence is considered. The anthropologist Frank Marlowe has suggested that hunter-gatherers with overnight camp groups, or bands, form into units of 30–50, group sizes that do not support Dunbar’s ideas. Dunbar has argued against this, pointing out that these bands are often unstable and are embedded in larger local communities.18 Comparison of human groups from hunter-gatherer bands to modern companies actually leads to an important and currently unanswered question: what sort of human social unit is appropriate to study if we want to draw inferences about human social evolution? Despite these criticisms, Dunbar asserts that ‘There is now considerable evidence that groupings of this size occur frequently in human social organisations, and that this is the normative limit on the size of personal social networks among adults.’ He is right about the evidence, but he was also a co-author on many of the studies he cites in support. That’s not a problem per se, but it would be good to have something even more solid on which to hang one’s scientific hat. Luckily, away from the vagaries and difficulties of social anthropology and the debates over the appropriate groups for meaningful study, there is now some solid neurological evidence supporting Dunbar. A number of studies have reported correlations between individual differences in social network size and the volumes of areas in the brain involved with social cognition, including the cortex and, intriguingly given its connections to stress (see Chapter 5), the amygdala.18
Dunbar’s number isn’t the only number in town. Some field studies in the United States led anthropologist H. Russell Bernard, Peter Killworth and colleagues to a mean average number of 290 (roughly double Dunbar’s) and a median for the now so-called Bernard-Killworth number of 231. The median is the ‘middle value’ in a distribution of data and the fact that this is lower than the mean indicates that some individuals had very large networks, pushing the mean average up. Although far less well known, the Bernard-Killworth number is nonetheless quite well supported using different methods, including asking people about their social networks and using statistical approaches to estimate people’s networks.19
Overall, despite some criticisms and a competitor in the shape of the Bernard-Killworth number, Dunbar’s number has pretty much survived the test of time. The details may have come under fire, but the notion that there is an upper limit on our ability to keep track of meaningful social interactions is now widely supported. What is more, set against our potentially vast online social networks, the exact value of this number is actually not that important as long as it remains, as it seems to be doing, in the range of a few hundred or so. Who cares if our upper limit is 150, 290 or even 500 when many of us are juggling active social networks that might well be in the 1,000 to 10,000 range?
We can get further insight into possible evolutionary constraints on our social networking through more detailed analyses that show that the relationships within the social network encapsulated by the Dunbar number are not the same. There are layers within the network that occur around 3 to 5, 9 to 15 and 30 to 45 individuals, with each successively larger layer incorporating individuals with progressively weaker ties, expressed both as perceived emotional closeness and frequency of interaction.20 These layers might explain the consistent and seemingly spontaneous formation of smaller social groups around these size bands, and the techniques used to identify these preferred group sizes can also account for layers beyond the Dunbar number, at 500 and 1,500 people. In these outer reaches of our social solar system, at 1,500 people, we reach the limit not of our cognitive ability to keep track of interactions but of something much more basic: our ability to remember that many faces.21
This layered analysis of social networks is neat and, like the overall theory, highly intuitive, but once again it can start to crumble around the edges when placed under pressure. Studies have shown that relationships simply aren’t tiered like an emotionally simplistic layer cake and that human social interactions are more complex and interconnected. Different relationships within the layers provide support for different things; thus a close family member might provide regular emotional support, while a friend in the next layer might provide occasional but exceptionally important financial support. In practice, there are multiple ways to define the ‘closeness’ that leads to people being placed in different layers.22
Studies have also attempted to define far larger networks than those estimated by the Dunbar number. One study in 2006 found that Americans know a mean of 610 and a median of 550 people. The range in network size was vast, with 90 per cent of the adult population knowing between 250 and 1,710 people, and half knowing 400 to 800 people. Another study, which followed people around in Malta, showed that people had networks of people that embed them in society, open up opportunities, provide useful services and give a sense of belonging that number around 1,000. To use these studies as arguments against the Dunbar number (as some have done22) is perhaps a little obtuse since Dunbar was not including these sorts of relationships in his definition, but overall it matters not. Whichever way we slice it, and despite some heated opposition, the conclusions actually turn out to be more or less the same: we have a limit in terms of social interactions that is imposed by the limitations of our brain, and that limit is likely in the order of hundreds rather than thousands of people. Even if we take the upper limits of those studies that do not support the Dunbar number we still max out around the 1,500 mark, which is a long way below the online social networks of a great many people, me included.
Is social media a mismatch?
If we accept that we have a constraint on the size of social networks in real life imposed by evolution, and that those constraints conspire to max us out at some level below what many people would consider to be their online network, then can we attribute any blame for the problems caused by social media to the mismatch between the evolved real and the created virtual world? I think the answer is a cautious yes.
The argument made by Dunbar in his 2012 analysis of social networking sites is that online networks were actually not especially large. Bearing in mind the date of the study, Twitter communities were found to be of the order of 100 to 200 (slap bang in Dunbar number territory) and ‘Facebook’s own data suggest that the claim that large numbers of friends is the norm is, at best, an exaggeration: while the distribution is certainly “fat-tailed” [some people have a lot of friends], the average and modal [the most common value] number of friends is in fact approximately 130’.18 What is interesting is that in 2019 the number of Facebook friends has only gone up to 155 and once again we find ourselves right in the Dunbar comfort zone. Twitter has slipped out of the zone though, with 707 followers on average in 2019, but even this value is still within the natural network range suggested by other researchers. Dunbar’s interpretation of larger group sizes on social media is either that such users are professionals using social media to reach audiences or that people are dragging ‘acquaintances’ from the 150 to 500 layer into their ‘friends’ group, in effect promoting people to friends but probably not interacting with them too much and certainly not feeling comfortable to sit down in a pub with them uninvited. The quality of those interactions is also often quite low-rent; a quick ‘like’ takes much less effort than a supportive conversation in real life.
The average size of online networks does not seem to indicate that we are in any sort of mismatch but such raw figures, useful though they are, do not tell the full story. For one thing, many users (the proportion of which will depend on the distribution of group sizes across all users) will have considerably more than the average-sized group and find themselves well outside of the comfort zone afforded even by the top end of normal social-group size estimates. Social networks online also exist separately from real-world networks, although with some overlap. Consequently, you might be close to your Dunbar number (or whatever estimate you choose to use) in real life and then easily max it out by having even a small network in the virtual world. We evolved for one world, and the Dunbar number and other models of social interaction do not really allow for two networks running side by side in different worlds with perhaps only limited crossover. I would argue that a great many people likely have a combined, complex social network that integrates real and virtual worlds (with some potentially difficult connections between them), the magnitude of which puts them well in excess of the social networks we have cognitively evolved to manage.
A simple count of friends or followers also greatly underestimates the connectivity and sheer neediness of online social networks. Depending on your privacy settings it is possible, indeed likely, for complete strangers to intrude on ‘your’ network, leave comments, butt in on conversations, challenge your opinions, insult you and even threaten you. These interactions, arising as a direct result of your social networking, would simply not be tolerated in the real world, but we are all still learning and creating the rules of engagement for the virtual world. Again, depending on your settings, each of these intrusions is a needy little notification requiring our attention, distracting us from activities and possibly our real-world network. Also, as we have seen, if such interactions are negative they can add greatly to the potential for anxiety, stress and depression, at least in some of us.
We have evolved to formulate and support small social networks that provide support and friendship and through that a series of selective advantages. We are a social species and we are stronger together. Over time we undoubtedly developed larger and larger social networks, but until recently such networks were not implicated in clinical conditions like depression or seen as sources of stress. Indeed, there is a strong body of research showing that friends and social interactions are overwhelmingly positive. In the last decade or so, though, we have created an online world that has become for some of us a major component of our real-world environment and which is perfectly suited for fostering very large, and potentially massive, networks. These networks may not require the level of individual attention that real-world friendships require, but collectively the scale of interactions that can be entered into, even with relatively modest social media networks spread across just a few platforms, can rapidly become an overriding obsession. Our brains simply aren’t allowing us to keep up. My advice: if it’s not working for you then don’t engage with it. Use it for what you want, and make sure you understand privacy and notification settings, because a constantly beeping mobile phone was definitely not part of our evolutionary background. Actually, that might make a good tweet; maybe it will go viral…
Notes
* The Conversation Fact Check is a part of The Conversation website. Specialists analyse stories from a wide range of topics including politics, science and medicine to get to the facts. As they say, they ‘ remove the spin from the debate ’. https://theconversation.com/uk/factcheck