Probability Waves

Double-slit experiments shattered the idea that particles just travel along single trajectories. Yet calling electrons "waves" raises an immediate question: a wave of what, exactly? The answer, developed by Max Born in 1926, is surprisingly elegant: the wave is a probability amplitude.

From intensity to probability

In ordinary optics, the intensity $I$ of a light wave is proportional to the square of the field amplitude $A$ : $I \propto A^2$ . Where the intensity is high, many photons land; where it is low, few do. This hints at the bridge between waves and probability.

For a matter particle, the same logic applies. Assign to every point in space and time a complex number $\Psi(\mathbf{r}, t)$ , called the probability amplitude or wavefunction. The quantity $|\Psi|^2$ then gives the probability density for finding the particle there:

P(\mathbf{r}, t)\,dV = |\Psi(\mathbf{r}, t)|^2\,dV.

This is the Born rule. Because $\Psi$ is complex, $|\Psi|^2 = \Psi^* \Psi$ where $\Psi^*$ is the complex conjugate. Crucially, the rule acts on the modulus squared, so the phase of $\Psi$ is not directly observable — yet it is physically real, because phases drive the interference that produces the diffraction pattern on the screen.

Why complex amplitudes?

A real-valued amplitude could only add constructively or destructively depending on sign. A complex amplitude $\Psi = |\Psi|\,e^{i\phi}$ carries a phase angle $\phi$ that can take any value in $[0, 2\pi)$ . When two paths contribute amplitudes $\Psi_1$ and $\Psi_2$ , the combined probability density is

|\Psi_1 + \Psi_2|^2 = |\Psi_1|^2 + |\Psi_2|^2 + 2\,\text{Re}(\Psi_1^*\,\Psi_2).

The cross-term $2\,\text{Re}(\Psi_1^*\,\Psi_2)$ is the interference term. It can be positive (constructive interference, bright fringe), negative (destructive interference, dark fringe), or zero (no interference; the intensities simply add) depending purely on the relative phase $\phi_1 - \phi_2$ . Remove the complex structure and you lose the ability to describe any of this.

Normalization

Because the particle must be somewhere, the total probability summed over all space must equal one:

\int_{\text{all space}} |\Psi(\mathbf{r}, t)|^2\,d^3r = 1.

This normalization condition is a constraint the wavefunction must satisfy. A solution of the wave equation that is not yet normalized can always be divided by the square root of its integral, provided that integral is finite, to produce a normalized wavefunction.

What the wave is NOT

It is tempting to picture $\Psi$ as a physical wave jiggling through space the way a water wave does. That picture fails for two reasons. First, for a system of $N$ particles, $\Psi$ lives in a $3N$ -dimensional configuration space, not ordinary 3-D space, so it cannot be "out there" in the room. Second, when a measurement is made, the particle is found at a single location; the probability density collapses, and the wavefunction changes discontinuously. No mechanical wave behaves this way.

The correct reading is epistemic: $\Psi$ encodes everything we can know about the particle's position before we look, and $|\Psi|^2$ tells us the distribution of outcomes we would see if we looked many times under identical preparation.