How Nature Works: The Science of Self-Organized Criticality
A book for general readers on self-organised criticality, which Bak proposed with Chao Tang and Kurt Wiesenfeld in 1987. Bak argues that complexity in nature reflects a tendency of large interacting systems, driven slowly, to organise themselves without outside tuning into a critical state where a minor disturbance may set off avalanches of all sizes. The emblem is a sandpile built grain by grain: once critical, one more grain may move a few grains or the whole pile, and in the model slide sizes follow a power law. He asks whether this is the single principle behind catastrophes that follow simple statistical laws, fractals, 1/f noise and Zipf's law, and calls it so far the only known general mechanism to generate complexity.
The book carries the idea to landscapes, earthquakes, solar flares, evolution, the brain, economics and traffic. In the evolution model he built with Kim Sneppen, the least fit species and its two neighbours repeatedly get new random fitnesses, and the system organises itself into punctuated equilibria, with avalanches of extinction events of all sizes; mass extinctions, Bak argues, could arise without a meteorite, though the theory cannot rule one out. If his picture holds, large catastrophes come from the same dynamics as small everyday events, so a specific explanation for each gives little insight, and their statistics can be predicted but not when the next will come. When the evolution model was replayed with the trigger of one large extinction removed, others came later, and in his sandpile, local barriers shrink avalanches only for a while.
Bak writes as co-originator and advocate. His models are deliberately oversimplified toys, and their reach across fields rests on universality, which he calls a hope. He concedes that some early sandpile experiments were not critical, presumably because of inertia, though an Oslo rice pile was; that the Game of Life's criticality is not robust and may be an accident; and that his evolution model is barely a skeleton for a theory of macroevolution. Levin (1998) called the critical-state claims for ecosystems stimulating but still without substantial empirical support. Bak touches on organisations only in passing; the bridge to psychological safety is ours. We read it against explaining each serious failure by its own cause: if large failures share the dynamics of small everyday ones, removing one trigger does not prevent the next, and, as Bak suggests for earthquakes, the many smaller events may give insight into the rare large ones.
Save this paper to a bibliography, keep notes on it, and export citations. Create a free account →