Home > Positioning > Persons > Bak

Per Bak (1948–2002)

Bak introduced self-organized criticality — the idea that many natural systems are driven toward critical states where small perturbations can trigger cascading events of any size, and that the system reaches this state on its own, without external tuning. The claim is bold: criticality is not a special condition that has to be arranged; it is the generic condition that extended dissipative systems evolve toward. The sand pile was the image; power-law distributions were the signature; the range of proposed applications — earthquakes, extinctions, financial crashes, neural activity — was deliberately provocative.

Life

Born 8 December 1948 in Brønderslev, Denmark. MSc in physics from the Technical University of Denmark; PhD in theoretical physics from the same institution (1974). Postdoctoral work at Cornell University. Research physicist at Brookhaven National Laboratory from 1977, where he spent the bulk of his career. Visiting positions at the Niels Bohr Institute and Imperial College London. Died 16 October 2002 in Copenhagen.

Self-organized criticality

The foundational paper: Bak, Tang, and Wiesenfeld, “Self-organized criticality: An explanation of the 1/f noise” (Physical Review Letters, 1987). The observation: many natural systems produce power-law distributions — the frequency of an event is inversely related to its size, with no characteristic scale. Earthquakes follow the Gutenberg-Richter law; species extinctions show similar scaling; electrical noise in many systems follows a 1/f spectrum. The question: why are power laws so common?

The answer Bak proposed: these systems have self-organized to a critical state — the boundary between ordered and disordered behaviour. At criticality, correlations extend across the entire system, and perturbations of any size can propagate. The system reaches this state through its own dynamics, without external tuning — hence “self-organized.” Classical statistical mechanics had studied criticality as a phenomenon that occurs at specific parameter values (temperature at a phase transition); Bak’s claim was that extended dissipative systems are attracted to criticality as a dynamical attractor.

The sand-pile model

The canonical illustration. Grains of sand are added one at a time to a pile. The pile builds until it reaches a critical slope. Beyond that slope, adding a single grain can trigger an avalanche — and the avalanche can be any size, from a single grain to a system-spanning cascade. The distribution of avalanche sizes follows a power law. No grain is special; the same local rule (add a grain, check stability, topple if unstable) produces cascades at all scales.

The sand-pile model is not a simulation of a literal sand pile — real sand piles show more complex behaviour. It is a minimal model that exhibits self-organized criticality: the system drives itself to a critical state and stays there, producing scale-free dynamics from purely local rules.

How Nature Works

How Nature Works: The Science of Self-Organized Criticality (Copernicus/Springer, 1996) is Bak’s synthesis. The book extends self-organized criticality from the sand pile to earthquakes (the Gutenberg-Richter law as a consequence of fault networks at criticality), biological evolution (extinction cascades as critical avalanches in ecosystems), economics (market crashes as critical phenomena), and neural activity (the brain operating near criticality to maximise its dynamic range).

The tone is characteristically assertive. Bak presents self-organized criticality not as one explanatory framework among others but as the fundamental mechanism behind a wide class of natural phenomena. The breadth of the claim is part of its appeal and part of what makes it contested.

Reception and critique

Self-organized criticality generated enormous interest and significant pushback. The interest: the framework offers a single mechanism for the ubiquity of power laws across domains, and the sand-pile model is simple enough to be analytically and computationally tractable. The pushback comes on several fronts.

Are the power laws real? Some proposed examples of self-organized criticality turn out, on closer analysis, to have characteristic scales or to follow distributions that are not strictly power-law. The statistical methods for identifying power laws in empirical data have been sharpened since Bak’s original claims, and some cases have not survived the scrutiny.

Is the mechanism generic? Bak claimed that extended dissipative systems generically evolve toward criticality. Critics argue that the conditions under which self-organized criticality arises are more specific than Bak suggested — that it requires particular kinds of driving and dissipation, not just any open system.

Does it explain or redescribe? The deepest critique: calling a power-law distribution “self-organized criticality” may name the phenomenon without explaining it. The sand-pile model shows that a specific rule set produces criticality; it does not show that the real systems Bak invokes operate by the same mechanism.

The concept has nonetheless become part of the standard vocabulary in statistical physics, geophysics, and neuroscience. The debate is about scope and specificity, not about whether self-organized criticality is a real phenomenon.

Where Bak stops

Bak’s programme gives a mechanism — local rules driving a system to criticality — and a signature — power-law distributions of event sizes. What it does not give is a theory of what happens at criticality beyond the statistics. The avalanche is described by its size distribution; its internal dynamics, its consequences for the system’s subsequent evolution, and its relationship to the system’s function (if any) are outside the framework. Self-organized criticality describes the regime; it does not describe what the regime produces.

Key works

See also: Kauffman · Complex Adaptive Systems