Thought Toys · Exhibit 18
Drop grains of sand one at a time. When a spot gets too steep it topples, spilling onto its neighbours — which can topple too. Most grains do nothing. Then one, no different from the rest, sets off an avalanche that crosses the whole table. The pile isn't tuned to do this; it tunes itself, to the very edge of stability.
—
—
Every cell is a little stack of sand. Add a grain and, if a stack reaches four, it topples: it hands one grain to each of its four neighbours and keeps none of the four. That can push a neighbour to four, so the topple spreads — an avalanche. Grains that fall off the table's edge are simply lost. That's the whole machine.
Start from an empty table and rain grains down. At first nothing avalanches — the pile is too flat, and almost every grain just sits. But the average slope creeps up on its own, and once it reaches a certain steepness the pile stops getting any steeper: every further grain that would pile up triggers a slide that carries the excess away. The pile has parked itself at a knife-edge — critical — without anyone setting it there. That's self-organized criticality.
At that edge, avalanches have no typical size. Watch the log–log plot fill in: the points lie on a straight line, which means a slide ten times bigger is just a fixed fraction rarer — there's no "normal" avalanche and no special "huge" one, only the same pattern at every scale. Most grains still do nothing; a rare grain takes down half the table; both come from the identical rule. Flip walls on and the edges stop draining: the table fills until it can't settle, and the scale-free behaviour dies. The same maths is blamed for the sizes of earthquakes, forest fires, and market crashes.
← the cabinet · Thought Toys — a cabinet of explorable explanations. Exhibit 18.