“There are techniques you can employ that allow you to foster desirable properties and reinforce positive behaviours.”
NICHOLAS CHRISTAKIS, Yale professor, on how to manipulate social networks.
N ever mind the selfies; we can manipulate social media to make the world a better place, says Dr. Nicholas Christakis. The professor of social and natural science at Yale University was speaking to Campaign on the sidelines of the recently held International Government Communication Forum in Sharjah, UAE.
Talk of influencers in social media tends to involve narcissistic 20-somethings with blogs about make-up. But clicks, likes and followers aren’t the only measure of influence. Users of social media can be studied scientifically to see how their actions affect others.
There are three primary ways that social media can be manipulated to “increase our health, our wealth, our creativity, our cooperation, our civic life and our security,” says Christakis. These are: connection, contagion and position.
Connection is how networks are structured, and who is connected to who. “It’s like the wiring pattern of a network,” says Christakis. “You could take a group of people and connect them one way, or you could take the same people and connect them in a different way.”
The second category is contagion. “Given a pattern of connections, given that I am connected to you and you to her, and whatever the connections are in a group,” he says, “Who should you target in that group to create an artificial tipping point? Who is that most structurally influential, or what is the most structurally influential subgroup of people?”
This subgroup, says Christakis, is that one that, “if you deliver an intervention to them, everyone copies them.” That is not unlike those tweeting millennials, but these users are found through their structural position rather than number of connections alone.
The third category of manipulation is position. “Position is that, given that you have a set of people and given that you have a particular structure, where should the people be put within the network?”
For example, he says: “Imagine you have a workplace and you have some people who are very energising and lift up their co-workers. Or imagine you have people who stimulate other people’s creativity. If you were a thoughtful manager you would think about not just the structure of your work groups but who you put where.”
Christakis’ insight comes from experiments carried out mainly in public health settings, and manipulating the structure of groups online.
Such changes are like restructuring carbon atoms. Connect them one way and you get soft, dark graphite. But assemble the same carbon atoms another way and you will get diamond: hard and clear.
“You could take a group of people and connect them one way, or you could take the same people and connect them in a different way.”
“There are two key intellectual ideas there,” he says. “First of all, these properties of softness and darkness and hardness and clearness aren’t properties of the carbon atoms; they are properties of the collection of carbon atoms. These are known as emergent properties.
“Second, which properties you get depends on how you connect the carbon atoms to each other. It’s the same with human groups. I can take a group of people and connect them in one way and they are very kind to each other, and very creative and healthy. Or I can take the same group of people and connect them in a different structure and they are none of those things: they are unkind to each other, or unhealthy or unproductive.”
In the public health field, manipulation of emergent properties might be applied to websites to link up people trying to quit smoking or to share information between diabetics about insulin shots, side effects and diet tips.
The people in those online networks can be organised so they reinforce each other’s good behaviour or so they reinforce one another’s bad behaviour.
“There are techniques you can deploy,” says Christakis: “connection, contagion and position – that allow you to foster desirable properties in this group, reinforce positive behaviours for example.”
The techniques are a doubleedged sword, though. “Dual-use technology,” Christakis calls them. Like nuclear power and guns, he says, they can be used for good or ill. As an example, he says, he has done research into using network interventions to increase voter turnout. But the same techniques could be used by states or bodies that want to suppress voter turnout.
“We have also done a lot of work on the spreading dynamics of truth and falsehood,” he adds. “Of course, we are interested in increasing the spread of true information, but other countries might be interested in increasing the spread of false information.”