Under some conditions, as group size increases the range of behaviors narrows.
Are there rules, akin to the laws of physics, that explain the patterns and regularities that arise in human society?
Participants in a three-day January workshop in Santa Fe, organized by SFI Postdoctoral Fellow Aaron Clauset and Michelle Girvan (former SFI postdoc, now University of Maryland), addressed that question drawing on principles of statistics, physics, computation, chemistry, political science, and sociology.
“The range of typical behaviors in society is narrower than one would expect,” says Aaron. “Embarrassment, peer pressure, and other social factors determine much of what I do.”
On a population-wide scale, these patterns often manifest themselves as predictable regularities, he says. Recent SFI work, for example, has found scaling laws for cities that suggest group-level behavior of residents of a community often is a function of the city’s size.
Much of the workshop discussion explored how social behavior at the individual level aggregates into more predictable group behavior, says Aaron, and how individual choice transitions into population behavior.
“As you increase group size, the range of behaviors narrows,” he says. “When do the laws of the group take over? It’s an important question.”
SFI External Professor Rob Axtell (Brookings Institution) described the use of two agent-based models to understand rm formation as a social phenomenon. One model depended on individual choice, and the other abstracted away individual behavior completely. Both methods produce the same results. In other words, says Aaron, some social phenomena can be viewed in two ways, one where individual choice matters, and one where it doesn’t.
David Gibson (University of Pennsylvania) discussed how the human desire to t in socially with others can shape the range of acceptable individual choices, in conversation or when standing in line, for example. “Thus, in some contexts, you might not even need that large a group before social constraints cause people to behave in highly predictable ways,” says Aaron.
Workshop participants also noted similarities and differences — methodologies, language, and level of investment, for example — between physics and sociology. One important difference is the relative difficulty in obtaining high-quality data in sociology, although computing and microelectronic technologies are increasingly benefitting sociologists, says Aaron.
Matthew Salganik (Princeton University) described research to understand fads in music, with results suggesting that the perception of the success of a pop culture product can become a self-fulfilling prophecy. The study is an example of large scale research made possible by the internet.
The workshop was successful because it brought together natural and social scientists, who, says Aaron, “had a productive and wide-ranging discussion about the scientific, cultural, and institu- tional factors impinging on our ability to answer the question of whether there is a physics of society.”
One outcome of the workshop, he says, might be a summary piece for Science or Nature, written by a small but interdisciplinary group of workshop participants, about the prospects and future of physics of society research.
For a complete list of workshop speak- ers and their presentation abstracts, visit SFI’s events page at www.santafe.edu/ events.