Connected: The Amazing Power of Social Networks and How They Shape Our Lives

Next, in our own work, we asked the respondents to tell us how interconnected their social contacts were to each other. So if a person said that Tom, Dick, Harry, and Sue were his friends, we asked him if Tom knew Dick, if Tom knew Harry, if Tom knew Sue, if Dick knew Harry, and so on. We then used these answers to calculate the probability that any two of a person’s friends were also friends with each other. This probability is an important property that we use to measure how tightly interwoven a network is.

If you know Alexi, and Alexi knows Lucas, and Lucas knows you, we say this relationship is transitive—the three people involved form a triangle. Some people live in the thick of many transitive relationships (like person A in the illustration on page 14), while others have friends who do not know each other (like person B). Those with high transitivity are usually deeply embedded within a single group, while those with low transitivity tend to make contact with people from several different groups who do not know one another, making them more likely to act as a bridge between different groups. Overall, we found that if you are a typical American, the probability that any two of your social contacts know each other is about 52 percent.

Although these measures characterize the networks we can see, they also tell us something about the networks we cannot see. In the vast fabric of humanity, each person is connected to his friends, family, coworkers, and neighbors, but these people are in turn connected to their friends, family, coworkers, and neighbors, and so on endlessly into the distance, until everyone on earth is connected (pretty much) to everyone else, one way or another. So whereas we think of our own network as having a more limited social and geographic reach, the networks that surround each of us are actually very widely interconnected.

It is this structural feature of networks that underlies the common expression “it’s a small world.” It is often possible, through a few connections from person to person, for an individual to discover a connection to someone else. A famous example (at least among social scientists) was described in a paper first drafted in the 1950s by two early figures in the study of social networks, Ithiel de Sola Pool and Manfred Kochen. One of the authors overheard a patient in a hospital in a small town in Illinois say to a Chinese patient in the adjoining bed: “You know, I’ve only known one Chinese before in my life. He was….from Shanghai.” Whereupon the response came back, “Why, that’s my uncle.”

(#litres_trial_promo) In fact, the authors did not tell us his name, perhaps because they were worried that the reader, in a further illustration of the small-world effect, would know him.

RULE 2: OUR NETWORK SHAPES US

Our place in the network affects us in turn. A person who has no friends has a very different life than one who has many. For example, we will see in chapter 4 (#litres_trial_promo) that having an extra friend may create all kinds of benefits for your health, even if this other person doesn’t actually do anything in particular for you.

One study of hundreds of thousands of Norwegian military conscripts provides a simple example of how the mere number of social contacts (here, siblings) can affect you.

(#litres_trial_promo) It has been known for some time that first-born children score a few points higher in terms of intelligence than second-born children, who in turn score a bit higher than third-born children. One of the outstanding questions in this area of investigation, however, has been whether these differences are due to biological factors fixed at birth or to social factors that come later. The study of Norwegian soldiers showed that simple features of social networks, such as family size and structure, are responsible for the differences. If you are a second-born son whose older sibling died while you were a child, your IQ increases and resembles the IQ of a first-born child. If you are a third-born child and one of your older siblings died, your IQ resembles that of a second-born child; and if both of your older siblings died, then your IQ resembles that of a first-born child.

Whether your friends and other social contacts are friends with one another is also crucial to your experience of life. Transitivity can affect everything from whether you find a sexual partner to whether you commit suicide. The effect of transitivity is easily appreciated by the example of how divorce affects a child. If a child’s parents are married (connected) then they probably talk to each other, but if they get divorced (disconnected) they probably do not. Divorce means that communication often has to pass through the child (“Tell your father not to bother picking you up next Saturday!”), and it is much harder to coordinate raising the child (“You mean your mother bought you ice cream too?”). What is remarkable is that even though the child is still deeply connected to both parents, her relationship with each of them changes as a consequence of the divorce. Yet these changes result from the loss of a connection between the parents—a connection the child has little to do with. The child still has two parents, but her life is different depending on whether or not they are connected.

And how many contacts your friends and family have is also relevant. When the people you are connected to become better connected, it reduces the number of hops you have to take from person to person to reach everyone else in the network. You become more central. Being more central makes you more susceptible to whatever is flowing within the network. For example, person C in the figure on page 14 is more central than person D. Ask yourself which person you would rather be if a hot piece of gossip were spreading; you should be person C. Now ask yourself which person you would rather be if a deadly germ were spreading in the network; you should be person D. And this is the case even though persons C and D each have the same number of social ties: they are each directly connected to just six people. In later chapters, we will show how your centrality affects everything from how much money you make to whether you will be happy.

RULE 3: OUR FRIENDS AFFECT US

The mere shape of the network around us is not all that matters, of course. What actually flows across the connections is also crucial. A bucket brigade is formed not to make a pretty line for you to look at while your house is burning but so that people can pass water to each other to douse the flames. And social networks are not just for water—they transport all kinds of things from one person to another.

As we will discuss in chapter 2 (#u8dbe58a0-6f74-5b9b-8d13-341fca80d633), one fundamental determinant of flow is the tendency of human beings to influence and copy one another. People typically have many direct ties to a wide variety of people, including parents and children, brothers and sisters, spouses (and nice ex-spouses), bosses and coworkers, and neighbors and friends. And each and every one of these ties offers opportunities to influence and be influenced. Students with studious roommates become more studious. Diners sitting next to heavy eaters eat more food. Homeowners with neighbors who garden wind up with manicured lawns. And this simple tendency for one person to influence another has tremendous consequences when we look beyond our immediate connections.

RULE 4: OUR FRIENDS’ FRIENDS’ FRIENDS AFFECT US

It turns out that people do not copy only their friends. They also copy their friends’ friends, and their friends’ friends’ friends. In the children’s game telephone, a message is passed along a line by each child whispering into the next child’s ear. The message each child receives contains all the errors introduced by the child sharing it as well as those introduced by prior children to whom the child is not directly connected. In this way, children can come to copy others to whom they are not directly tied. Similarly, every parent warns children not to put money in their mouths: the money, we think, contains germs from numerous people whose hands it has passed through, and not just from the most recent pair of hands. Analogously, our friends and family can influence us to do things, like gain weight or show up at the polls. But their friends and family can influence us too. This is an illustration of hyperdyadic spread, or the tendency of effects to spread from person to person to person, beyond an individual’s direct social ties. Corto’s brother lost his life because of such spread.

It is easy to think about hyperdyadic effects when the network is a straight line—(“that guy three people down the line better pass the bucket, or we’re all going to be in big trouble”). But how on earth can they be understood in a natural social network such as the college students in the illustration on page 14, or complex networks of thousands of people with all kinds of crosscutting paths stretching far beyond the social horizon (as we will consider later)? To decipher what is going on, we need two kinds of information. First, we must look beyond simple, sequential dyads: we need to know about individuals and their friends, their friends’ friends, their friends’ friends’ friends, and so on. And we can only get this information by observing the whole network at once. It has just recently become possible to do this on a large scale. Second, if we want to observe how things flow from person to person to person, then we need information about the ties and the people they connect at more than one point in time, otherwise we have no hope of understanding the dynamic properties of the network. It would be like trying to learn the rules of an unfamiliar sport by looking at a single snapshot of a game.

We will consider many examples and varieties of hyperdyadic spread, but we can set the stage with a simple one. The usual way we think about contagion is that if one person has something and comes into contact with another person, that contact is enough for the second person to get it. You can become infected with a germ (the most straightforward example) or with a piece of gossip or information (a less obvious example). Once you get infected by a single person, additional contact with others is generally redundant. For example, if you have been told accurately that stock XYZ closed at $50, another person telling you the same thing does not add much. And you can pass this information on to someone else all by yourself.

But some things—like norms and behaviors—might not spread this way. They might require a more complex process that involves reinforcement by multiple social contacts. If so, then a network arranged as a simple line, like a bucket brigade, might not support transmission of more complicated phenomena. If we wanted to get people to quit smoking, we would not arrange them in a line and get the first one to quit and tell him to pass it on. Rather, we would surround a smoker with multiple nonsmokers, perhaps in a squad.

Psychologist Stanley Milgram’s famous sidewalk experiment illustrates the importance of reinforcement from multiple people.

(#litres_trial_promo) On two cold winter afternoons in New York City in 1968, Milgram observed the behavior of 1,424 pedestrians as they walked along a fifty-foot length of street. He positioned “stimulus crowds,” ranging in size from one to fifteen research assistants, on the sidewalk. On cue, these artificial crowds would stop and look up at a window on the sixth floor of a nearby building for precisely one minute. There was nothing interesting in the window, just another guy working for Milgram. The results were filmed, and assistants later counted the number of people who stopped or looked where the stimulus crowd was looking. While 4 percent of the pedestrians stopped alongside a “crowd” composed of a single individual looking up, 40 percent stopped when there were fifteen people in the stimulus crowd. Evidently, the decisions of passersby to copy a behavior were influenced by the size of the crowd exhibiting it.

An even larger percentage of pedestrians copied the behavior incompletely: they looked up in the direction of the stimulus crowd’s gaze but did not stop. While one person influenced 42 percent of passersby to look up, 86 percent of the passersby looked up if fifteen people were looking up. More interesting than this difference, however, was that a stimulus crowd of five people was able to induce almost as many passersby to look up as fifteen people did. That is, in this setting, crowds larger than five did not have much more of an effect on the actions of passing individuals.

RULE 5: THE NETWORK HAS A LIFE OF ITS OWN

Social networks can have properties and functions that are neither controlled nor even perceived by the people within them. These properties can be understood only by studying the whole group and its structure, not by studying isolated individuals. Simple examples include traffic jams and stampedes. You cannot understand a traffic jam by interrogating one person fuming at the wheel of his car, even though his immobile automobile contributes to the problem. Complex examples include the notion of culture, or, as we shall see, the fact that groups of interconnected people can exhibit complicated, shared behaviors without explicit coordination or awareness.

Many of the simple examples can be understood best if we completely ignore the will and cognition of the individuals involved and treat people as if they were “zero-intelligence agents.” Consider the human waves at sporting events that first gained worldwide notice during the 1986 World Cup in Mexico. In this phenomenon, originally called La Ola (“the wave”), sequential groups of spectators leap to their feet and raise their arms, then quickly drop back to a seated position. The effect is quite dramatic. A group of physicists who usually study waves on the surface of liquids were sufficiently intrigued that they decided to study a collection of filmed examples of La Ola in enormous soccer stadiums; they noticed that these waves usually rolled in a clockwise direction and consistently moved at a speed of twenty “seats per second.”

(#litres_trial_promo)

To understand how such human waves start and propagate, the scientists employed mathematical models of excitable media that are ordinarily used to understand inanimate phenomena such as the spread of a fire through a forest or the spread of an electrical signal through cardiac muscle. An excitable medium is one that flips from one state to another (like a tree that is either on fire or not) depending on what others around it are doing (are nearby trees on fire?). And these models yielded accurate predictions of the social phenomenon, suggesting that La Ola could be understood even if we knew nothing about the biology or psychology of humans. Indeed, the wave cannot be understood by studying the actions of a single individual standing up and sitting down. It is not orchestrated by someone with a megaphone atop a cooler. It has a life of its own.

Mathematical models of flocks of birds and schools of fish and swarms of insects that move in unison demonstrate the same point: there is no central control of the movement of the group, but the group manifests a kind of collective intelligence that helps all within it to flee or deter predators. This behavior does not reside within individual creatures but, rather, is a property of groups. Examination of flocks of birds “deciding” where to fly reveals that they move in a way that accounts for the intentions of all the birds, and, even more important, the direction of movement is usually the best choice for the flock. Each bird contributes a bit, and the flock’s collective choice is better than an individual bird’s would be.

(#litres_trial_promo) Similar to La Ola and to flocking birds, social networks obey rules of their own, rules that are distinct from the people who form them. But now, people are not having fun in a stadium: they are donating organs or gaining weight or feeling happy.

In this regard, we say that social networks have emergent properties. Emergent properties are new attributes of a whole that arise from the interaction and interconnection of the parts. The idea of emergence can be understood with an analogy: A cake has a taste not found in any one of its ingredients. Nor is its taste simply the average of the ingredients’ flavors—something, say, halfway between flour and eggs. It is much more than that. The taste of a cake transcends the simple sum of its ingredients. Likewise, understanding social networks allows us to understand how indeed, in the case of humans, the whole comes to be greater than the sum of its parts.

Six Degrees of Separation and Three Degrees of Influence

Stanley Milgram masterminded another, much more famous experiment showing that people are all connected to one another by an average of “six degrees of separation” (your friend is one degree from you, your friend’s friend is two degrees, and so on). Milgram’s experiment, conducted in the 1960s, involved giving a few hundred people who lived in Nebraska a letter addressed to a businessman in Boston, more than a thousand miles away.

(#litres_trial_promo) They were asked to send the letter to somebody they knew personally. The goal was to get it to someone they thought would be more likely than they to have a personal relationship with the Boston businessman. And the number of hops from person to person that the letter took to reach the target was tracked. On average, six hops were required. This amazing fact initiated a whole set of investigations into the small-world effect originally characterized by de Sola Pool and Kochen, and it entered popular culture too, with John Guare’s play Six Degrees of Separation and even the trivia game Six Degrees of Kevin Bacon.

But some academics were skeptical. For instance, as far apart as Nebraska and Boston might be (both geographically and culturally), they were both inside the United States. So in 2002, physicist-turned-sociologist Duncan Watts and his colleagues Peter Dodds and Roby Muhamad decided to replicate Milgram’s experiment on a global scale using e-mail as the mode by which people communicated.

(#litres_trial_promo) They recruited more than ninety-eight thousand subjects (mostly from the United States) to send a message to “targets” around the world by forwarding the e-mail to someone each subject knew who might in turn know the targeted person. Each subject was randomly assigned one target from a list of eighteen possible targets in thirteen countries. The targets included a professor at an Ivy League university, an archival inspector in Estonia, a technology consultant in India, a policeman in Australia, and a veterinarian in the Norwegian army—quite a motley crew. Once again—astonishingly—it took roughly six steps (on average) to get the e-mail to each targeted person, replicating Milgram’s original estimate of just how small the world is.

However, just because we are connected to everyone else by six degrees of separation does not mean that we hold sway over all of these people at any social distance away from us. Our own research has shown that the spread of influence in social networks obeys what we call the Three Degrees of Influence Rule. Everything we do or say tends to ripple through our network, having an impact on our friends (one degree), our friends’ friends (two degrees), and even our friends’ friends’ friends (three degrees). Our influence gradually dissipates and ceases to have a noticeable effect on people beyond the social frontier that lies at three degrees of separation. Likewise, we are influenced by friends within three degrees but generally not by those beyond.

The Three Degrees Rule applies to a broad range of attitudes, feelings, and behaviors, and it applies to the spread of phenomena as diverse as political views, weight gain, and happiness. Other scholars have documented that among networks of inventors, innovative ideas seem to diffuse to three degrees, so that an inventor’s creativity influences his colleagues, his colleagues’ colleagues, and his colleagues’ colleagues’ colleagues. And word-of-mouth recommendations for everyday concerns (like how to find a good piano teacher or how to find a home for a pet) tend to spread three degrees too.

There are three possible reasons our influence is limited. First, like little waves spreading out from a stone dropped into a still pond, the influence we have on others may eventually peter out. The stone displaces a certain volume of water as it is dropped, and the energy in the wave dissipates as it spreads out. One way to think about this socially is that there is decay in the fidelity of information as it is transmitted, as in the child’s game of telephone. So, if you quit smoking or endorse a particular political candidate, by the time this information reaches your friends’ friends’ friends’ friend, that person may no longer have accurate or reliable information about what you actually did. We call this the intrinsic-decay explanation.

Second, influence may decline because of an unavoidable evolution in the network that makes the links beyond three degrees unstable. Ties in networks do not last forever. Friends stop being friends. Neighbors move. Spouses divorce. People die. The only way to lose a direct connection to someone you know is if the tie between you disappears. But for a person three degrees removed from you, any of three ties could be cut and you would lose at least one pathway between you. Hence, on average, we may not have stable ties to people at four degrees of separation given the constant turnover in ties all along the way. Consequently, we do not influence nor are we influenced by people at four degrees and beyond. We call this the network-instability explanation.

Third, evolutionary biology may play a part. As we will discuss in chapter 7 (#litres_trial_promo), humans appear to have evolved in small groups in which everyone would have been connected to everyone else by three degrees or less. It is indeed useful to know whether anyone in our group has it in for us or is our ally, or whether others need our help or might help us. And it is useful to influence others in our group to do what we do. But we have not lived in large groups long enough for evolution to have favored people who can extend their influence beyond three degrees. Put another way, we may not be able to influence people four degrees removed from us because, in our hominid past, there was no one who was four degrees removed from us. We call this the evolutionary-purpose explanation.

It seems likely that all these factors play a role. But no matter the reasons, the Three Degrees Rule appears to be an important part of the way human social networks function, and it may continue to constrain our ability to connect, even though technology gives us access to so many more people.

While this inherent limit may seem, well, limiting (who doesn’t want to rule the world?), we should remember how small the world is. If we are connected to everyone else by six degrees and we can influence them up to three degrees, then one way to think about ourselves is that each of us can reach about halfway to everyone else on the planet.

Moreover, even when restricted to three degrees, the extent of our effect on others is extraordinary. The way natural social networks are structured means that most of us are connected to thousands of people. For example, suppose you have twenty social contacts, including five friends, five coworkers, and ten family members, and each of them in turn has similar numbers of friends and family (to make things simple, let’s assume they are not the same contacts as yours). That means you are indirectly connected to four hundred people at two degrees of separation. And your influence doesn’t stop there; it goes one more step to the twenty friends and family of each of those people, yielding a total of 20 ? 20 ? 20 people, or eight thousand people who are three degrees removed from you. That would include every single person in the small Oklahoma town where James grew up.

So while the observation that there are six degrees of separation between any two people applies to how connected we are, the observation that there are three degrees of influence applies to how contagious we are. These properties, connection and contagion, are the structure and function of social networks. They are the anatomy and physiology of the human superorganism.

Connected

Most of us are already aware of the direct effect we have on our friends and family; our actions can make them happy or sad, healthy or sick, even rich or poor. But we rarely consider that everything we think, feel, do, or say can spread far beyond the people we know. Conversely, our friends and family serve as conduits for us to be influenced by hundreds or even thousands of other people. In a kind of social chain reaction, we can be deeply affected by events we do not witness that happen to people we do not know. It is as if we can feel the pulse of the social world around us and respond to its persistent rhythms. As part of a social network, we transcend ourselves, for good or ill, and become a part of something much larger. We are connected.

Our connectedness carries with it radical implications for the way we understand the human condition. Social networks have value precisely because they can help us to achieve what we could not achieve on our own. In the next few chapters, we will show how networks influence the spread of joy, the search for sexual partners, the maintenance of health, the functioning of markets, and the struggle for democracy. Yet, social-network effects are not always positive. Depression, obesity, sexually transmitted diseases, financial panic, violence, and even suicide also spread. Social networks, it turns out, tend to magnify whatever they are seeded with.

Partly for this reason, social networks are creative. And what these networks create does not belong to any one individual—it is shared by all those in the network. In this way, a social network is like a commonly owned forest: we all stand to benefit from it, but we also must work together to ensure it remains healthy and productive. This means that social networks require tending, by individuals, by groups, and by institutions. While social networks are fundamentally and distinctively human, and ubiquitous, they should not be taken for granted.

If you are happier or richer or healthier than others, it may have a lot to do with where you happen to be in the network, even if you cannot discern your own location. And it may have a lot to do with the overall structure of the network, even if you cannot control that structure at all. And in some cases, the process feeds back to the network itself. A person with many friends may become rich and then attract even more friends. This rich-get-richer dynamic means social networks can dramatically reinforce two different kinds of inequality in our society: situational inequality (some are better off socio-economically) and positional inequality (some are better off in terms of where they are located in the network).

Lawmakers have not yet considered the consequences of positional inequality. Still, understanding the way we are connected is an essential step in creating a more just society and in implementing public policies affecting everything from public health to the economy. We might be better off vaccinating centrally located individuals rather than weak individuals. We might be better off persuading friends of smokers of the dangers of smoking rather than targeting smokers. We might be better off helping interconnected groups of people to avoid criminal behavior rather than preventing or punishing crimes one at a time.