Mathematics of Patterns

Lesson 3 of 4 · Intermediate · 18 min read

The Golden Ratio

The golden ratio, denoted by the Greek letter phi, is an irrational number approximately equal to 1.6180339887. It arises when a line is divided into two parts such that the ratio of the whole line to the longer part equals the ratio of the longer part to the shorter part.

a / b = (a + b) / a = φ ≈ 1.618

This proportion appears with striking frequency in nature, art, and architecture. The Parthenon's facade, Leonardo da Vinci's Vitruvian Man, and the spiral of a nautilus shell all exhibit proportions close to the golden ratio. Whether this reflects a deep mathematical truth about aesthetics or a tendency to find the ratio where we look for it remains a subject of scholarly debate.

The golden ratio is intimately connected to the Fibonacci sequence. As Fibonacci numbers grow larger, the ratio between consecutive terms converges on phi. This connection explains why Fibonacci numbers appear so frequently in natural growth patterns — they are mathematical approximations of golden-ratio proportions.

a b a + b φ
The golden rectangle and its inscribed spiral

Symmetry Groups

Symmetry is the mathematical study of transformations that leave a pattern unchanged. There are four fundamental types of symmetry operations:

In 1891, the Russian crystallographer Evgraf Fedorov proved that there are exactly 17 distinct wallpaper groups — 17 fundamentally different ways to repeat a pattern across a flat plane using combinations of these symmetry operations. Remarkably, all 17 groups can be found in the decorative art of the Alhambra palace in Granada, Spain, created centuries before the mathematics was formalised.

Reflection Rotation Translation
The three fundamental symmetry operations

Tiling Mathematics

A tiling (or tessellation) covers a flat surface completely using one or more shapes with no gaps and no overlaps. The mathematics of tiling asks fundamental questions: which shapes can tile the plane? In how many ways? And can a set of tiles force a non-repeating pattern?

Regular tessellations use a single regular polygon. Only three regular polygons can tile the plane on their own: equilateral triangles, squares, and regular hexagons. When combinations of regular polygons are allowed, eight additional semi-regular (Archimedean) tessellations become possible.

Penrose tiles: In the 1970s, Roger Penrose discovered sets of just two tile shapes that can cover the plane but only in a non-periodic way — the pattern never exactly repeats. These aperiodic tilings exhibit five-fold symmetry, something impossible in periodic tilings, and have deep connections to quasicrystals discovered in the 1980s.
Left: periodic hexagonal tiling. Right: aperiodic Penrose-style tiling

Chaos Theory & Fractals

Chaos theory studies deterministic systems whose behaviour is highly sensitive to initial conditions — the famous "butterfly effect." Despite being governed by precise rules, chaotic systems produce behaviour that appears random and unpredictable. Yet within this apparent randomness, beautiful patterns emerge.

Fractals are the geometric language of chaos. A fractal is a shape that exhibits self-similarity: each part resembles the whole at a different scale. The term was coined by Benoit Mandelbrot in 1975. Unlike the smooth shapes of classical geometry (lines, circles, spheres), fractals have fractional dimensions and infinite detail.

The Mandelbrot set, generated by iterating the simple equation z = z² + c in the complex plane, produces an infinitely intricate boundary that reveals new patterns at every magnification. Julia sets, Sierpinski triangles, Koch snowflakes, and Menger sponges are other well-known fractals, each arising from simple recursive rules.

The Sierpinski triangle — a fractal built from recursive subdivision
Key insight: Fractals bridge the gap between mathematics and nature. Coastlines, mountain profiles, cloud boundaries, and tree canopies all have fractal properties. Mandelbrot famously asked: "How long is the coast of Britain?" — a question with no single answer, because the measured length depends on the scale of measurement.