Inventing the Social Network
What did people do before they networked?
Maybe they pulled strings, used connections, kept good company, took advantage of protectzia, or deployed guanxi. “Networking,” as in “making use of a network of people,” did not even appear in English until the late 1970s. “Network” itself usually meant a television broadcasting system such as NBC; the notion of a social network was essentially unknown. Today, most Americans understand a social network as a set of Internet contacts—and perhaps more subtly, per the movie of that name, as a comment on Facebook co-founder Mark Zuckerberg, who may not have gotten so far without his connections.
The concept of a social network emerged from sociologists’ and anthropologists’ efforts in the 1960s and early 1970s to unpack conventional understandings of group and community and to understand how individuals operate through strategic personal connections. Nancy Howell Lee explained how, in an era when abortions were illegal, pregnant women nonetheless found abortionists: by asking someone who knew someone who knew someone else. Mark Granovetter showed how professionals in the Boston area got better jobs: very often through “weak ties” such as former classmates passing on tips during passing encounters. Charles Kadushin described how becoming a psychiatric patient in New York involved traveling in the circles of psychiatric aficionados. Elihu Katz and colleagues revealed how media affect most consumers’ decisions: not directly, but through the influence of friends and families who more closely follow the media.
Even earlier, British anthropologists, especially in Africa, found that supplementing old categories such as tribe and lineage with the concept of social network—defined as a specific set of interpersonal linkages that help to explain individuals’ behavior—revealed dynamics that had been previously hidden. These anthropologists showed how rural migrants to the booming cities settled in and thrived far from home by using old connections and building new ones. In the 1970s scholars such as Barry Wellman explained that the classic “community question”—what has happened to community in the modern world?—rests on the mistaken notion that community is a fixed place and argued that a community is really a web of personal ties that can transcend space.
Other scholars soon found the social network perspective useful. Political scientists, for example, showed how neighbors and friends influenced voters’ choices. Health researchers reported that people with strong ties, who have “social support,” live longer. In a striking case, economists Dora Costa and Matthew Kahn found that the loyalty and survival rates of soldiers in the Civil War, including those interned as prisoners, were greater if the troops served with men from their hometowns.
Policymakers also began to rely on the network concept. In the early 1980s, for instance, the California Department of Mental Health launched the “Friends Can Be Good Medicine” campaign, telling Californians that if they set aside their shyness and bonded more closely, they would live longer. The program clearly sought to replace government spending with the labor of friends and family, but the ideological commitment to smaller government came with some basis. Networks often sustain individuals by providing practical help and advice, modeling good behavior while shaming bad, making people feel appreciated, and even through direct physiological responses to company. Health workers in developing countries are mobilizing networks to spread family planning, AIDS prevention, childhood immunization, and other practices. Hearing about, say, the protective effects of male circumcision from a childhood buddy is more effective than listening to radio appeals or speeches by officials.
Judith Levine found that welfare reform in the 1990s led the poor mothers she studied both to turn to their networks far more and to face far more demands for help from their networks, which may well have been what welfare critics wanted. But it strained already stressed social ties.
Other interested policymakers track terrorist networks. The recent controversy over NSA surveillance reveals the degree to which security agencies have bought into the power of network analysis. They want to know the contacts of the contacts of suspects, to display those networks visually, and to examine them using the latest math tools.
Network thinking helps us better understand issues like psychological stress and community dynamics.
From the start, the network metaphor lent itself to mathematical realization. A latticework of links among nodes—a network of relationships among people—is a graph just like an electrical grid is. Innumerable calculations follow. For example, the ratio of actual to possible connections among a set of people is “density,” an indicator of how tightly knit that group is. The number of connections going through a particular person indicates his or her “centrality” and thus clout.
In the popular domain, networks caught on as a way of describing how people actively pursue and employ relationships than can be helpful: networking, the verb. Before 1980, 38 books in the Library of Congress had “networking” in their titles; in the 1980s, 146 new ones did; and in the 1990s, publishers put out 690 networking books. Some of the upsurge probably reflects the spreading popularity of the term in fields from neuroscience to home Internet installation, but many of these books focused on human networks—in business and marketing, politics, and general schmoozing. The metaphor had escaped from the lab. Massive popularization arrived, of course, with social media—first MySpace and Friendster, later Facebook, Google+, and so on.
Network thinking has empowered researchers, change-makers, and the general public. It helps us better understand issues such as psychological stress and community dynamics. We see how connections extend the inequality produced by structural and political forces. Network analysis has enabled researchers to weigh more precisely and see more exactly how the conventional wisdom of “It’s who you know” plays out—in which circumstances and for which people.
As with any new idea, network analysis can be oversold, both as a research tool and a source of policy. The boldest claims for network effects appeared several years ago in a series of studies by Nicholas Christakis and James Fowler, who professed to find that people were likelier to be smokers, overweight, or unhappy if the associates of their associates were—even if associates in the middle were not. Much debate ensued. Even network enthusiasts wondered about influence at two or more steps removed.
Friends and relatives do influence behavior, but individuals also pick which friends and which relatives to spend time with based on behavior. Do teens drink because their friends drink or do teens who drink become friends? Both. Network theory doesn’t provide a complete explanation. Trying to disentangle and weigh the two processes is difficult. Much research suggests that the selection effect—choosing or being chosen by people like you—is usually more important than the influence effect.
Although networks are widely celebrated and their impact now confirmed, one must acknowledge that friends and family are not just good medicine but also stressors. Often the same people can have both effects. To amplify Sartre, l’enfer, comme le ciel. Les autres require attention and help; they find fault; they can lead us into bad habits (such as teen drinking) and cause worry. A few studies find that the more people one cares about, the more emotionally taxed one is by those people’s difficulties. (Among my people, illnesses are often explained as the consequence of aggravation and that aggravation as the consequence of imperfect children.) Some research suggests that whether networks help us physically and mentally is less important than believing they do.
Personal networks are most critical, scholars generally believe, in societies where markets and governments do not effectively deliver. However much networking we do in the United States, our efforts are nothing like those in countries with less effective markets and governments—a classic example is Soviets trying to find out when consumer goods would be available and how to reach the person distributing them. Still, Americans do network—as they did before they knew they were networking. Only now, thanks to the metaphor from the academic lab, we rely on considerably more technology and tutoring, and do so with far greater self-consciousness.
Image: Matthew Powell