The story
In 1963, the meteorologist Edward Lorenz of the Massachusetts Institute of Technology published a paper, 'Deterministic Nonperiodic Flow', showing that even a very simple weather model can be impossible to predict over the long run. His model had just three equations, run on an early computer. Small differences in the starting values grew rapidly, so two almost identical starts produced completely different results. This is the core of what came to be called chaos.
The discovery came in the winter of 1961. In the common telling, Lorenz restarted a simulation from the middle by typing in numbers printed to three decimal places, when the machine had used six. The run quickly drifted away from the original, and he realized the tiny rounding difference was the cause. Newton's laws had suggested a clockwork universe in which perfect knowledge of the present gives a perfect forecast. The French mathematician Henri Poincare had sensed in the 1890s that some systems escape this, but Lorenz showed it on a computer, in a model of heated air in convection.
The idea took years to spread. Plotted over time, the Lorenz system traces a looping shape, the 'Lorenz attractor', that never repeats yet stays within bounds. The phrase 'butterfly effect' comes from a 1972 talk with the title 'Does the Flap of a Butterfly's Wings in Brazil Set off a Tornado in Texas?' The mathematicians Tien-Yien Li and James Yorke named the field 'chaos' in 1975. It placed a hard limit on weather forecasting, which loses skill after roughly two weeks, and led forecasters to run many simulations from slightly different starts to measure uncertainty.
Why it mattered
- Weather and climate science adopted ensemble forecasting, running many simulations to estimate how uncertain a forecast is.
- Chaos theory spread across physics, biology and economics, studying systems from dripping taps to heart rhythms to populations.
- It helped inspire fractal geometry and the study of complex systems in the 1970s and 1980s.
- The idea reshaped scientific ideas about prediction, showing that determinism does not guarantee predictability.
Sources
- When the butterfly effect took flight MIT Technology Review
- Edward Norton Lorenz MacTutor, University of St Andrews
- Born half a century ago, chaos theory languished for years Science News
Drafted by AI (Claude) from the sources above and checked against them by AI; no historian has reviewed this page. Spot a mistake? Suggest a correction. How pages are made
Community notes
Loading notes…