Far from the Ivy-covered walls that gave rise to Facebook, the first social network analysis took place amid some far higher and more restrictive barricades—the ones that surround Sing Sing prison.

In 1931 my father, the psychiatrist and psychodrama founder J.L. Moreno, was approached by a national prison committee to help reorganize a model prison, Sing Sing. Within a few months, J.L. was assigning these hardened thugs and swindlers roles to play in theater games to assess their spontaneity, which he called group therapy. Then he and his team reassigned them to work groups with leaders based on their natural choices, a system that was later named sociometrics. 

Within a few months J.L. was given another assignment, to address an outbreak of girls running away from an upstate New York reform school. Writing on BuzzFeed in 2012, tech journalist Russell Brandom observed that “The Future of Facebook Was Born in 1932” with the elaborate social graphics my dad created to reorganize the New York State Training School for Girls:

Fourteen girls had run away in two weeks. Instead of examining each case individually, he mapped all fourteen girls on a graph, showing how each case socially influenced the others, eventually leading to a social kind of epidemic. As you may have noticed, it looks an awful lot like Facebook’s social graph. … And for anyone skeptical about the social graph, it’s a powerful reminder: people have been at this for quite a while, and modern social networks are barely scratching the surface.

Between 1928 and 1933, J.L. (who died in 1974 at the age of 84), conducted the first social network experiments at Brooklyn public and private schools, Sing Sing prison, and what was then called a reformatory for delinquent girls. Though he later became famous (relatively speaking, of course) as the founder of psychodrama, a form of psychotherapy that uses roleplaying, his were the first attempts to graph interpersonal relations in real life in what he called sociograms, the key devices of his new science of sociometry. As I explain in my new book, Impromptu Man: J.L. Moreno and the Origins of Psychodrama, Encounter Culture, and the Social Network, my dad’s intuition about the invisible links between us became the platform for modern social media.

The Italian computer scientist Massimo Franceschetti writes that for all the popularity of social networks today, “few know, however, that the first example of social network is attributed to psychiatrist Jacob Moreno, a hand-drawn image depicting friendship patterns between the boys and the girls in a class of schoolchildren, presented at a medical conference in New York in 1933.” At P.S. 181 in Brooklyn, my dad and his team observed relationships in a kindergarten class and then in classes through the eighth grade. The children’s interpersonal choices were recorded in “sociograms” based on whom they chose to sit next to while studying or playing. Triangles were used to signify boys and circles for girls, with their initials in the middle.

One look at the sociogram of the fourth grade shows that not much has changed in 80 years, at least not in kid society. The boys are clustered on one side, the girls on the other, with only one precocious young man reaching out to one girl. There is also an isolated pair of girls, which might mean trouble in class or on the playground or even out of school.   If you were the teacher of this class today, this sociogram might cause you to keep an eye out for bullying, in school or in cyberspace.

The fact that there is an implicit structure within any group creates tremendous opportunities for research and, as we now know, for making a lot of money through modern social media that takes advantage of the Internet. For many years, gathering data and drawing sociograms was cumbersome and labor-intensive. When I was a kid, our living room floor was often covered with large sheets of paper and students working with colored pencils as they navigated and updated the islands of social graphs, and that was only after doing the analysis of choices within and between groups. As Web entrepreneurs realized in the 1990s, inter-personal choices could be collected instantly and charted in free software. 

Through social media, a new era of social network analysis has been born, and so have new controversies. A few months ago Facebook took a lot of criticism for its study of “emotional contagion” when it manipulated users’ positive and negative posts. These make for interesting academic exercises, but J.L. would surely have objected to the concentration of power and knowledge about social networks by a few giant social media companies. He was never interested in data collection for its own sake, and certainly not only as a business model. As a matter of principle he believed that the results should always be the basis for intervening in the life of the group to improve the lives of its members. He called his sociometry the science of, by, and for the people. His enthusiasm for social reform, his lack of training in modern statistical methods, and bombastic manner made him an outlier in the world of academic social science after World War II. For the last 30 years of his life J.L. focused on psychotherapy rather than social science.

But it is hard to overstate the importance of his pioneering projects. The authors of a 2009 paper in the journal Science began their review of network analysis social science with a description of J.L.’s experiments in the 1930s. They are not only the ancestors of social media like Facebook and Twitter but also for organizational assessments in government, industry, the military, and civil society. “Network research is ‘hot’ today,” the authors of the Science paper wrote, and because of tools made possible by the Internet it’s getting hotter: Three times as many papers were published on “social networks” in 2009 as in 1999.

Every time I see a social graphic—from my Facebook friends network to the National Security Agency’s graphs of electronic contacts to studies of obesity as predictable based on social networks—I think of J.L. scrawling those first little sociograms more than 80 years ago. What would he make of modern social media? I’m confident he would say that they should not only be used to peer into our lives but to make them better. Will those who follow us in another 80 years say that they did?