Fractals
The Geometry That Nature Actually Uses
Every tree, coastline, and cloud is built from a shape that Euclid never imagined — one that stays equally complex no matter how far you zoom in.
The Idea
Classical geometry is a set of idealisations. Spheres, cubes, straight lines — these are useful fictions, tools for approximation. But when Benoit Mandelbrot looked at the natural world in the 1970s, he noticed something embarrassing: almost nothing in nature actually resembles these shapes. A mountain is not a cone. A cloud is not a sphere. A coastline is not a line at all. Mandelbrot's insight was to take this messiness seriously rather than smooth it away. The shapes he was interested in share a property called self-similarity: they look roughly the same at every scale. Zoom into a fern frond and you find smaller frond-shapes. Zoom into those and you find smaller ones still. The pattern recurs, not perfectly, but recognisably, across orders of magnitude. What makes this mathematically striking is that fractal objects have a dimension that isn't a whole number. A smooth line has dimension 1. A flat plane has dimension 2. But a fractal coastline — crinkled, detailed, endlessly complex — has a dimension somewhere between 1 and 2, perhaps 1.25. This fractional dimension is a precise measure of how aggressively a shape fills space. The real revelation isn't just aesthetic. Fractal geometry turns out to describe how biological systems maximise surface area within constrained volumes — which is exactly the engineering problem your lungs, your blood vessels, and your brain have all solved.
In the World
In 1967, Mandelbrot published a paper with a deceptively simple title: 'How Long Is the Coast of Britain?' His answer: it depends entirely on the length of your ruler. Measure the coastline with a 100-kilometre ruler and you skip over every bay and peninsula, getting a relatively tidy number. Switch to a 10-kilometre ruler and you trace more of the detail — the number grows. Use a 1-kilometre ruler and it grows further still. There is no convergence. The more closely you look, the longer the coast becomes, in principle without limit. Different countries had been measuring their shared borders with different-sized 'rulers' and genuinely getting different answers — which explained a persistent, embarrassing discrepancy in international maps. This wasn't a measurement problem. It was a geometry problem. The coastline isn't a line in any conventional sense; it's a fractal, with a measured dimension of around 1.25. The same logic was later applied to the bronchial tree in human lungs. Your airways branch and rebranch roughly 23 times, following a fractal pattern that packs an internal surface area of around 70 square metres — something close to the floor area of a modest apartment — into a space you could hold in two cupped hands. Engineers designing artificial lungs or drug-delivery systems now use fractal branching models directly. Mandelbrot's coastal surveying problem turned out to be the same problem as breathing.
Why It Matters
Most of us were taught geometry as a set of rules about perfect shapes, then sent out into a world that contains almost none of them. Fractal geometry is the corrective — a reminder that precision and messiness aren't opposites, and that the 'noise' we habitually filter out is often where the real structure lives. This has a broader epistemic implication. When we smooth data, round edges, or look for clean lines in complex phenomena — whether in economics, ecology, or medicine — we risk discarding the very information that carries meaning. The roughness isn't a flaw in the data; it may be the signal. On a more immediate level, recognising self-similarity trains a particular kind of attention. You start noticing that the branching of rivers from above looks like the branching of capillaries from below, or that the way a crack propagates through glass follows the same logic as a lightning bolt. These aren't coincidences — they're the same underlying mathematics expressing itself across wildly different physical systems. Once you see it, the world becomes demonstrably stranger and more coherent at the same time.
A Question to Ponder
If the most efficient structures in nature — lungs, blood vessels, neural networks — are fractal, what does that suggest about the kinds of problems fractal geometry might still be waiting to solve?
Get a new one of these every morning.
Start learning with Thinkable