Innovation and discovery are perhaps social phenomena, but that has not kept physical scientists from studying them. Human progress often appears to come out of nowhere, but the causative factors are anything but isolated. One thing leads to another; discoveries create conditions that lead to more discoveries, albeit in varying degrees.
Innovations, of course, occur in more general settings as well; the theory of punctuated equilibrium tells us that evolution of all species happens in fits and starts. Physicists have studied such bursts of change with mean-field methods, which led to the understanding that a phase transition separates systems showing intense activity from those where activity always dies out.
In their paper in Physical Review Letters, Vishal Sood of the Niels Bohr Institute, Denmark, Amer Shreim, Peter Grassberger, and Maya Paczuski at the University of Calgary, Canada, and Myléne Mathieu at the Ecole Normale Supérieure de Lyon, France, take us a step further with the help of a microscopic model of innovation that builds on earlier so-called branching process schemes. Their particular model—the interacting branching process—indicates, in contrast to earlier studies, that a pair of inventions (parents) is needed to generate an “offspring” innovation. Rather than associating periods of high innovation with a phase transition, the model tells us that fluctuations can lead to a “superexplosive” burst after a long, quiescent bottleneck. How closely this explains explosive sociological or historical changes is anyone’s guess, but it is tempting to think that there might be an explanation in physics of how we emerged from the Dark Ages into Renaissance, or why the World Wide Web arrived with breakneck speed. – Sami Mitra