Social Physics; how good ideas can be reviewed

Throughout this book I have argued that we need to think about society as a network of individual interactions rather than as markets or classes. To accomplish this, I have presented a social physics framework that outlines how the flow of ideas from person to person shapes the norms, productivity, and creative output of our companies, cities, and societies.

Alex Pentland, Social Physics


Alex Pentland’s Social Physics puts the single most important paragraph at the end of the book. In fact, the most valuable parts of the book are in the second half—and barely rely on the first half of the book for their clarity. Despite this, and my visceral reaction Pentland’s non-stop self-aggrandizement, the book has a lot of valuable components. In particular, the social physics framework Pentland creates is at the very minimum an enlightening way to analyze human interactions. At best, Pentland and his team at the MIT Media Lab are beginning to revolutionize our understanding of sociology, economics, and human psychology.  In other words, it is worth at least a cursory review, and an alternative reading order.


I. What does the book say

The foundation of Pentland’s theory is that human interaction is defined by idea flow, “The propagation of behaviors and beliefs through a social network by means of social learning and social pressure.” In general, the more idea flow the better. He contrasts this approach to the atomistic approach generally espoused by neoclassical economics; what matters for productivity is not the intelligence of individuals but rather the collective intelligence of a group as produced in large part by its rates of idea flow. He presents a variety of studies to prove this point, all relying on huge datasets he and his colleagues at the MIT Media Lab have analyzed.

Most notably, he worked with Bank of America’s call centers to run an experiment on idea flow. The call center he was looking at tried to minimize socialization between workers and, to that end, coffee breaks were given to one worker at a time. Because this was a call center of over three thousand employees, it was just as easy to give breaks to groups of twenty at a time, which Pentland’s group did to increase idea flow. This resulted in a decrease in average call handle time—the major cost of the call center—and caused Bank of America to change all of their call centers to this break system and “forecast a $15 million per year productivity increase.”

Idea flow is the product of a pair of processes, exploration and engagement. In short, exploration brings new ideas into a group and engagement is the process by which those ideas are integrated into the group’s behavior. The social physics framework thus looks at how to increase rates of communication among group members and with individuals outside of the group. In other words Pentland wants to build cohesive groups, rather than isolated individuals, and he wants to prevent those groups from becoming echo chambers.

To further generalize this theory and make it applicable, Pentland argues that engagement and exploration generate idea flow through a combination of social learning and social pressure which mean, conveniently, more or less what it sounds like they mean. Social learning is either learning new strategies or new beliefs through experience or observing other’s behavior. Social pressure is a fancy way to say peer pressure; it is the influence an individual exerts on another based on their relationship and interactions. Social learning is in large part the product of engagement and exploration while social pressure is the process by which social learning is integrated into a group and, in doing so, increase group intelligence (which he finds to be more predictive of productivity than the talent of the individuals within the group).


II. What does the book teach 

Something Pentland largely fails to do is explicitly integrate the components of his theory for the reader. This is a real shame because the social physics approach he outlines, while not the “new science” he claims it to be, is a valuable framework, so I’m going to try to summarize it.


Humans are social animals whose actions are heavily influenced by those they interact with and the relationships those interactions build. These interactions circulate ideas and norms both within peer groups and outside of them, and much like mixing genes this mixing of ideas serves to increase the diversity of thought within a society. Extending the metaphor, idea flow is as necessary for the health of a society as gene mixing is for the survival of a species; without new ideas entering a group it will become an echo chamber and without internal interactions any ideas entering cannot spread, nor can norms and relationships of trust be established.  Therefore a society is best off when it has a diversity of groups which interact with one another in a fluid fashion to maximize idea flow on the whole.

Put together, the social physics framework provides a model of an ideal society and a series of causal mechanisms that both explain a society’s health and can be analyzed (quantitatively, no less). How useful the theory proves to be is a question I would like to explore at a later date, as well as some of its parallels to work Pentland doesn’t mention, especially from psychology and sociology.

Leave a Reply

Scroll to top
Insert math as
Additional settings
Formula color
Text color
Type math using LaTeX
Nothing to preview