Born: how a scattering direction became a probability

The wave spreads out, but each detection is localized

Repeat the same collision many times: send an electron beam at a target and count the particles that come out in each direction. A wave calculation gives a spread-out pattern. The experiment gives individual, localized detections that build up over many repetitions. Linking the two descriptions needs a rule.

Bound-state energies alone did not supply that rule. Getting the hydrogen spectrum right still left open what the wavefunction stood for. In his 1926 collision work, Born used scattering and target excitation to ask what the mathematical amplitudes predict about distinct physical outcomes.

Read the evidence: M. Born (1926), Quantenmechanik der Stoßvorgänge, Z. Phys. 38, 803–827 — English translation of the extended collision paper

First choose the measurement, then square the amplitude

For a normalized pure state and an ideal measurement in an orthonormal basis, the modern expression below gives the probability of each outcome. The bra labels the outcome, the ket labels the prepared state, and the sum runs over a complete set. The complex amplitude itself is not something you count.

This compact form is more general and more standard than the language of the 1926 paper. In a scattering experiment, turning an outgoing wave into a count rate also needs the incident flux, the outgoing flux and the detector acceptance. Inelastic channels can send particles out at different speeds, and simply squaring a coefficient can miss that weighting.

p_n=\left|\langle n\mid\psi\rangle\right|^2,\qquad \sum_n p_n=1

Read the evidence: M. Born (1926), Quantenmechanik der Stoßvorgänge, Z. Phys. 38, 803–827 — English translation of the extended collision paper M. Born (1954), The Statistical Interpretation of Quantum Mechanics — Nobel lecture, December 11

Target excitation adds a second label

An electron can leave at a given angle with the target unchanged, or with the target raised by an allowed excitation energy. The outgoing direction and the target’s final state together define a channel. Energy conservation limits which channels are open, but it does not pick the outcome of any single collision.

This links the discrete energy losses seen in electron-impact experiments to a theory of how often each outcome occurs. The Franck–Hertz evidence for excitation thresholds and Born’s reading of channel amplitudes are two complementary steps, not two names for the same experiment.

Interference survives in the statistics

When two indistinguishable alternatives lead coherently to the same recorded outcome, their amplitudes must be combined first and only then turned into a probability. Treating them as an ordinary mixture generally gives a different prediction. The statistics keep information about relative phase, even though the detector reports only individual events.

A useful test therefore keeps the preparation fixed and changes only whether the alternatives stay coherent or become distinguishable. A smooth histogram on its own is not enough, because classical waves and classical random processes also produce smooth distributions. The controlled comparison is what tells the cases apart.

Read the evidence: M. Born (1954), The Statistical Interpretation of Quantum Mechanics — Nobel lecture, December 11

The Born rule and the Born approximation are different things

The Born approximation is a method for estimating scattering amplitudes when conditions allow. The Born rule says how amplitudes relate to outcomes. If the first approximation fails, the scattering calculation needs improving; that failure does not show the probability rule is wrong.

Likewise, running a random sampler with the rule built in shows the rule’s statistical consequences; it does not derive the rule from experiment. A real test has to account for background counts, efficiency, drift in the preparation and the uncertainty that comes from a finite number of observations.

Look at the apparatus before debating interpretation

The qubit Lab shows how outcomes depend on the measurement basis, and the partial-wave Lab shows angular distributions in a central-potential model. Neither is Born’s historical collision apparatus. The states and measurements each one assumes are part of the model you are exploring.

Matching outcome statistics tests a theory’s predictions. On its own it does not settle every philosophical question about the wavefunction or explain each individual event. Born’s historical contribution is clearest when that predictive achievement stays in view.