Only connect! That was the whole of her sermon.
Only connect the prose and the passion, and both will be exalted,
And human love will be seen at its height.
Live in fragments no longer.
Only connect…

This was the exhortation of E.M. Forster who, in Howards End (1924), explored the the difficulties in, and promises of, connecting in Victorian society prior to the first world war. Eighty five years later Hilary Clinton in a speech on internet freedom argued for enshrining Forster’s epigram to Howards End in the charter for universal human rights: ‘only connect’ becomes the right to connect.

According to Nicholas Christakis and James Fowler in Connected: The Amazing Power of Social Networks and How they Shape our Lives, the connections that bind form the bedrock of what it means to be human. If Forster was exhorting and Clinton demanding, then Christakis and Fowler are claiming that both miss the point: we cannot – and never have been able to – not connect. Connecting is both in our DNA and a characteristic of our pro-social intelligence. What is interesting is not that we connect but how we connect, the results of the connections and how an understanding of both can lead to a theory which has both explanatory and predictive power.

That theory builds on the idea that the sum can be greater than the parts – and has echoes of the hive mind and swarm intelligence so inimical to Jaron Lanier. Christakis and Fowler are however, at pains to detail the principles that govern contagion and connection in order to explain why social networks exist and how they work. These principles are:

1. we shape our network 2. our network shapes us 3. our friends affect us 4. our friends’ friends’ friends affect us 5. the network has a life of its own

So we can still choose the connections that we make and how salient each connection might be in a particular network but once that network is established, including our position in it,  it can shape our behaviour in different ways. It can make us happy, rich, poor and suicidal. That our friends influence our behaviour seems obvious but that our friends’ friends’ friends – individuals that we may never know – can influence us seems less so. Until we begin to think of social norms and the attempts to influence them. The reinforcement of complex processes involves multiple social contacts – so it’s argued, surrounding a smoker by multiple non-smokers is better than getting one ex-smoker to persuade a smoker to quit. Because the network has emergent properties (a cake is not reducible to an average taste of each of its ingredients, it transcends those ingredients and emerges as something else) the connections that are present can lead to something that transcends those connections. So, the argument implies, get the network ingredients right and we could cook goodness and happiness. We could of course also cook evil and although the authors spend some time looking at Milgram’s experiments with obedience to authority, they could have explored in much greater depth the ways in which evil can be an emergent property of particular social engineering. By coincidence, as I was reading Connected I also watched The Wave, an example of the ways in which group conformity can be manipulated through the ties that bind, leading to predictably frightening consequences. Evil is an emergent property but I’m not convinced that it is exclusively emergent in particular networks.

But to return to what Christiakis and Fowler did do. The notion that ideas and behaviour can be infectious and even contagious is not a new one. In the 1940s and 50s social scientists were studying how behaviour flows through groups. Paul Lazerfeld, in looking at the ways in which products became popular, saw that there was a two stage process involved: product popularity starts with an inner group and moves through connections outwards. So developed the idea of ‘viral’ marketing – creating the buzz with the cool kids, the influencers, at the centre and with the idea that those at the outer reaches of the network will become infected. The medicalisation metaphors continued in the 1990s with ‘epidemics’ of obesity, suicide, and more recently loneliness. Much of this earlier work was reconstructed after the event. Some behaviour became prevalent and the researchers tried to explain it. Only more recently with the digitisation of everyday behaviour through social networking sites have social scientist been able to look at dynamic networks and map their changes in shape, constitution as well as the issues that travel through them. What Christiakis and Fowler have been able to do however is to take data from an extensive ongoing study of heart disease (the Framingham Heart Study in the US which has followed more than 15,000 people, over 50 years and three generations) which has collected information on social connections between the people involved and begin to map those connections against the spread of various behaviours. They found that as people are connected so is their health connected.

There are clearly some unanswered questions with the research they conducted. In the UK there is a north-south divide in life expectancy with the more affluent regions recording higher levels of longevity than the poorest. The question is: are people living shorter lives in Glasgow because they are connected to people living shorter lives? Or are there other factors involved? Christakis and Fowler look for their answers in the network but surely environmental factors are playing a large part. Of course, some behaviours are contagious – even behaviours we enact ourselves. Charlie Chaplin suggested you should ‘smile when your heart is breaking ‘ because simply by smiling ‘you’ll find that life is still worthwhile’. Or perhaps you’ll find other smiley people and as ‘birds of a feather flock together’ you’ll find yourself in the company of like-minded people. This idea – homophily – is a powerful explanation of many of the ways in which we behave in groups and in opposition to other groups. Homophily is the glue that binds so much of our networks and which creates the kinds of self-prophesying ideas and echo-chambers increasingly being witnessed in our digital connections.

There are also a couple of dangerous implications buried in Connected. If I know my health and happiness are tied to my connections, why not simply break the unhealthy connections and stick with the healthy ones? Cut those ties to smokers … even to friends who have friends who are smokers. Want to improve the health of a particular network? In the same way as viral marketing works best from the most connected cool kid to the least connected kid on the margins, so better health starts with the most connected ‘node’ in the network and spreads out to those on the margin. Logically, health care benefits should be given to those people first and then it’ll spread out – good health will be a contagion as obesity is. The policy implications are not lost on Christakis and Fowler who chillingly suggest in a sentence that echoes Minority Report:

Correlating people’s network centrality with their mortality risk, their transitivity with their prospects of repaying a loan, or their network position with their propensity to commit crimes or quit smoking offers new avenues for political intervention.