We are rewriting the story of human populations with data that depict individual people simultaneously from above and below: at scales geographic and genetic, from social networks to microbial networks. What is new is not the aspiration to integrate each level of human experience, but the data that make it possible.
When we have only one kind of data, we can find only one kind of answer. But in the social sciences, explanations are like ecosystems. The presence of people—and their leftovers—enables mice to live in a house; the presence of a cat constrains them. Just as a species’ niche expands or contracts with the presence of other species, explanatory factors, too, are constrained and enabled by the presence of other factors. Data that combine disparate scales reveal this expansion and contraction of explanatory space.
Consider what makes someone smoke tobacco. We all know that, in the United States, fewer people smoke today than fifty years ago, just before a major cultural and regulatory shift began. What the sociologist Jason Boardman and his colleagues have now shown is that whether someone smokes is more heavily influenced by their genes today than it was before smoking was stigmatized. In the 1960s, when every hostess had an ashtray and every stranger had a light, it didn’t take much to decide to light up; today, when nicotine comes with dirty looks, it often takes a powerful biological urge that afflicts us unequally. The changed culture makes room for our genes to determine whether we smoke; our genes limit the room for the cultural shift to change what we do. Data only on genes or only on the shift in norms would give us one kind of answer about why people smoke, but both together show us how each constrains the other.
Or consider the rise of antibiotic-resistant staph bacteria. The epidemiologist Diane Lauderdale and her colleagues are analyzing a particular cause of this deadly epidemic in Chicago: prison. Their work triangulates knowledge at the micro scale of how the bacterium passes from one person’s skin to another’s with knowledge at the macro scale that determines whose skin touches whose: how people move in and out of crowded jails, where they live when they leave, what sports they play. The result is increasingly realistic models of interaction between microbes and humans—not only as individuals or as populations, but both at the same time.
This is the future of the population sciences: zooming simultaneously inside individuals, to their microbes or their genes, and outside of them, to their social norms, their neighborhoods, the laws that regulate them. Data that zoom in both directions don’t just let us ask new questions—they let us ask a new type of question, one that embraces the contingent and contextual nature of human behavior. The social sciences vacillate between broad generalities (that usually turn out to be less general than they appear) and particularistic studies of specific settings. Only data that link the levels of human experience let us fill in the gloriously contingent middle. In the human sciences, the scope conditions are the story.
When you’re making a map, you don’t just want to know what goes inside the borders; you want to know where the borders are. Explanations should map the space of possibilities, and data that span the levels of human variation let us explore the borders.