PER BAK 1948–2002
Credit and © "The Paula Gordon Show"
Per Bak died on October 16, 2002, at the age of 54. Per was one of the founders and most influential contributors to the study of complex systems. Per made many contributions to science, but the best known was a general theory of self-organization, which he called, "self-organized criticality". His ideas and discoveries have had an influence over how people think about a broad range of phenomena, from physics to biology, neurosciences, cosmology, earth sciences, economics and beyond. As a scientist and as a person he was an inspiration and a challenge to those of us who knew him. He was, for me that rare scientist who, though not in my field, could at any moment surprise me by saying something that would make me realize I had to rethink something I had thought I understood.
Per's theory of self-organized criticality was formulated in a paper he wrote in 1987 with two younger colleagues, Chao Tang and Kurt Wiesenfeld. This was one of the most highly cited and influential physics papers of the last two decades. It presented a general mechanism by which systems which are out of thermal equilibrium may evolve to a fractal, or scale invariant, distribution. These distributions are characteristic of many non-equilibrium systems, but before their paper no one had understood why. This idea, and the methodology it spawned, has been applied to understand the patterns that form spontaneously in many different systems, including earthquakes, forest fires, traffic jams, economic markets, and biological phenomena ranging from natural selection to the distribution of species of trees in a forest. It has also been applied to astrophysical phenomena ranging from x ray busts to the distribution of galaxies in the universe. Many of these applications were pioneered by Per himself, in collaboration with specialists in these fields.
Per's career proves that a scientist can indeed be a public intellectual, for the influence of his work extends far beyond science. Al Gore mentioned self-organized criticality in his book, "Earth in the Balance". I've heard executives of powerful software companies say they did not understand why their business strategies succeeded until they read Per's popular book, How Nature Works. At a meeting in Santa Fe I heard Lt. General (Retired) Paul Van Riper, one of the planners of the Gulf War, say that he learned so much from that book that could be applied to military strategy that he made it required reading for all U.S. Marine officers in training.
There was no one better than Per at grasping the heart of a complicated phenomena, and then realizing his insight by the invention of a simple model. Per's models were so elegant in their elimination of extraneous features that some experts were unable to understand why they applied at all to their subject. You could explain any of his models to a child, but to understand how they worked you had to be able to be able to strip everything away from a phenomena but the mechanisms by which a pattern forms and propagates. A tree on fire is like a person with the flu, is like a galaxy forming stars, is like a gene turning on, is like a trader making a bid...and each is represented by a single bit: all that matters is how many neighbors have to be on for it to light up. I remember him telling me once how excited he had been when he had accidentally run into Stephen J Gould. He introduced himself and very excitedly said that he had invented a theory that could explain Gould and Eldridges notion of evolution by punctuated equilibrium. Gould replied that he wasnt interested because "punctuated equilibrium is already a theory". Per was very disappointed. Several years later I was fortunate myself to meet Gould and I was glad to be able to report back to Per that I had spent a dinner explaining his theory to Gould and he had finally got it.
Spending time with Per was an exhilarating but sometimes daunting experience. Per's energy exceeded that of most people, he was full of ideas, and completely direct in expressing his opinions. If he liked an idea he was childlike in his enthusiasm; he praised it unreasonably, no matter whether it was his or someone elses. But if he thought something was stupid, he said so bluntly, no matter his relation with the speaker, or the consequences.
Per was perhaps the most fearless scientist I ever met. He more than once began a talk at a conference of neuroscientists or cosmologists with the statement that after six months of thinking about their field, he realized that everything done in the last 20 years was wrong. Having said that he would explain where he saw the mistake and propose a simple theory that reversed it, which he illustrated by a simple computer model. One such talk I heard, about how the brain works, was titled, "Learning from mistakes." Certainly some people went away unhappy, but I am convinced this was more childlike simplicity than arrogance. It would not have occurred to him that there was any other way to be: science is hard and we have to say what we think.
What Per understood better than any other scientist Ive known is that doing good science takes courage, stubbornness and a willingness to take big risks in the hope of making big advances. He argued stubbornly when he felt that someone did not understand one of his ideas, but he did not seem to mind if one of his theories was shown wrong or improved upon.
What Per certainly had no patience for was anything that smacked of insincerity or hypocrisy. If there was not a good reason to do something, he didnt waste his time, nor did he see why anyone else should. He was the kind of professor that university administrators most fear: a renowned scientist who simply said no when asked to take time and energy away from science in order to do something that interested him less at that moment.
The philosopher Roberto Mangabeira Unger explains some of the problems faced by contemporary scientists by saying that "we are something relatively infinite caught within finite realities: the body, society and culture." Certainly this is how I remember Per.