The Uncertainty Principle

Core idea

You cannot, even in principle, know both where a particle is and how fast it is moving with arbitrary precision. The product of the two uncertainties has a hard floor set by Planck's constant. This is not a limit of our measuring instruments; it is a property of the world. Once Heisenberg made it explicit in 1926, classical determinism — Laplace's dream of computing the entire future from a snapshot of the present — was dead.

Hawking's argument: Quantum mechanics introduces an unavoidable element of randomness into physics. It predicts probabilities, not outcomes, and that is the deepest reason classical mechanics had to be replaced.

Why it matters

Why classical physics had to break

By 1900 a quiet catastrophe had been brewing: hot bodies, classically calculated, should radiate infinite total energy across all frequencies — the so-called ultraviolet catastrophe. Planck patched the maths by assuming radiation comes in discrete packets, quanta, whose energy is proportional to frequency. At high frequencies the energy per quantum becomes prohibitive, so the radiation curve falls back to zero. Planck thought of his quanta as a calculation trick. Einstein, explaining the photoelectric effect five years later, took them literally — light really comes in packets, photons.

Heisenberg's bound

To measure where a particle is, you have to bounce something off it. To resolve a position better, you need a shorter-wavelength probe, which carries more energy per quantum. The collision then kicks the particle harder, scrambling its momentum. Heisenberg formalised this: Δx · Δp ≥ ℏ/2. The relation does not depend on the apparatus — it falls out of the wave-like structure of quantum states. A particle does not, before measurement, have simultaneously definite position and velocity to learn.

Wave–particle duality and the double slit

Light, classically a wave, behaves like a particle when emitted or absorbed (quanta). Particles, classically little dots, behave like waves when propagating: send single electrons one at a time through two slits and an interference pattern still builds up on the screen — each electron, in some sense, goes through both slits. The mathematics of quantum mechanics gives every conceivable path between two events a complex amplitude, and observable probabilities come from summing those amplitudes (Feynman's sum over histories). Particles and waves are not the underlying reality; they are two language tools we choose between depending on the question we ask.

Why the deterministic universe is gone

Determinism required two ingredients: laws that propagate states forward in time, and a precise present state to start from. Quantum mechanics keeps the first but kills the second. The Schrödinger equation is deterministic about the evolution of the wave function, but the wave function itself only encodes probabilities for measurement outcomes. The collapse from probabilities to a single observed value is the irreducibly random part — the part Einstein refused to accept (his "God does not play dice").

Where general relativity meets its limit

Quantum effects are negligible for planets and galaxies but dominant near the singularities general relativity itself predicts: the big bang and the centres of black holes. Where gravity gets strong enough to compress matter to atomic scales, the two great twentieth-century theories meet and contradict each other. The rest of the book is largely about that collision.

Key takeaways

Mental model

Copy sourceZoom

Practical application

For practical reading: quantum mechanics underwrites essentially every electronic device. The behaviour of electrons in semiconductors, the operation of LEDs, the way MRI scanners image hydrogen nuclei — all of this requires quantum mechanics. The uncertainty principle is not a curiosity tucked away in textbooks; it is the constraint that allows quantum tunnelling to occur, and tunnelling is what makes flash memory work. Every smartphone uses a phenomenon Einstein found unacceptable.

Example

Consider how data persists in NAND flash memory. A floating-gate transistor traps electrons inside an insulating barrier; the trapped charge encodes a bit. To write or erase, electrons must cross the barrier — but classically they have nowhere near enough energy to do so. They get through by quantum tunnelling, which exists precisely because position-momentum uncertainty allows an electron's wave function to extend through a classically forbidden region with non-zero probability. The probability is tiny per attempt, but with billions of electrons and trillions of attempts per second, it is enough to write a bit reliably.

A related example: scanning tunnelling microscopes image individual atoms by riding the same tunnelling current. The current is exponentially sensitive to the tip–surface gap, which is why STM can resolve atoms — and that exponential sensitivity comes straight out of the wave function's exponential decay in classically forbidden regions, itself a direct consequence of the uncertainty relation.

Related lessons