the Heisenberg uncertainty principle
/ HYE-zen-berg /
You cannot simultaneously pin down both where a particle is and how fast it is going. This is not a failing of your apparatus or your skill; it is a structural feature of the quantum world. The Heisenberg uncertainty principle says that certain pairs of properties are fundamentally incompatible — sharpening one inevitably blurs the other — because a state cannot be an eigenstate of both at once.
For position and momentum the statement is sigma_x sigma_p >= hbar/2, where sigma is the standard deviation (the spread) of measurement results over an identically prepared ensemble. The general version, the Robertson relation, ties any two observables to their commutator: sigma_A sigma_B >= (1/2)|<[A, B]>|, so the position-momentum bound is just [x, p] = i hbar in disguise. The deep reason is Fourier: a wavefunction narrow in position is necessarily broad in momentum, because position and momentum wavefunctions are Fourier transforms of each other, and no function can be sharply localized in both a variable and its transform.
Two honest clarifications. First, the principle is a statement about the spreads intrinsic to a quantum state — a property of how the state is prepared — not primarily about a measurement disturbing the system (Heisenberg's original microscope heuristic conflated the two; the disturbance version is a separate, subtler statement). Second, the energy-time relation Delta E Delta t >= hbar/2 has a different character, since time is a parameter rather than an operator; there Delta t means a characteristic timescale over which the state changes appreciably, not the uncertainty of a time observable.
Confining an electron to an atom-sized region Delta x about 1e-10 m forces a momentum spread Delta p >= hbar/(2 Delta x), giving a kinetic energy of order electron-volts — this zero-point motion is exactly what keeps the electron from spiraling into the nucleus.
Uncertainty is not a nuisance but the reason atoms are stable and have a size.
It bounds the intrinsic spreads of a prepared state (via the commutator), and is not the same statement as measurement disturbing the system; the energy-time relation is different again because time is not an operator.