Probability Amplitudes in Historical Context

Quantum mechanics did not merely attach probabilities to otherwise classical alternatives. It introduced a new intermediate object: the probability amplitude. Amplitudes can be complex, carry phase, add coherently, and cancel. Probabilities appear only after the relevant amplitudes have been combined for a specified experimental question.

The formal definition lives in Probability Amplitudes. This page explains why amplitude-based reasoning became historically unavoidable and how it connected wave mechanics, scattering, interference, path integrals, and later quantum field theory.

Amplitudes Versus Probabilities

A classical probability is a real number between $0$ and $1$. Mutually exclusive alternatives add:

$$P(A\ \text{or}\ B) = P(A)+P(B)$$

when $A$ and $B$ cannot both occur.

A classical wave uses a different rule. Fields add first, and intensities are quadratic in the resulting field. Quantum mechanics keeps the wave-like addition rule but applies it to amplitudes for detection outcomes. The result is neither an ordinary particle ensemble nor an ordinary material wave.

For a state expanded in an orthonormal basis,

$$\lvert\psi\rangle = \sum_n c_n\lvert n\rangle,$$

the coefficient $c_n$ is an amplitude for outcome $n$ when the measurement is in that basis. The probability is not $c_n$ itself. It is

$$P(n) = \lvert c_n\rvert^2.$$

The same idea appears in position space:

$$\psi(x) = \langle x\vert\psi\rangle, \qquad \rho(x) = \lvert\psi(x)\rvert^2.$$

The wavefunction supplies amplitudes; its squared modulus supplies a probability density. This distinction was not obvious at the birth of wave mechanics. Schrödinger initially explored more directly wave-like readings of $\psi$. Born’s scattering work made the probabilistic reading decisive: the wavefunction is not a literal charge density for a single electron, but a tool for assigning probabilities to outcomes.

Interference Terms

The signature of amplitude reasoning is the rule “add amplitudes first.” If two coherent alternatives contribute amplitudes $A_1$ and $A_2$ to the same outcome, then

$$P = \lvert A_1+A_2\rvert^2.$$

Expanding gives

$$\begin{aligned} P = {}& \lvert A_1\rvert^2 + \lvert A_2\rvert^2 \\ &+ 2\operatorname{Re}(A_1^*A_2). \end{aligned}$$

The last term is the interference term. It depends on relative phase. If $A_2=A_1e^{i\phi}$, then

$$P = \lvert A_1\rvert^2 \lvert 1+e^{i\phi}\rvert^2 = 2\lvert A_1\rvert^2(1+\cos\phi).$$

The probability can be enhanced, suppressed, or even vanish by destructive interference. A theory that only assigns probabilities to pre-existing alternatives cannot reproduce this rule without adding additional structure.

The contrast with classical alternatives is sharp:

Situation	Rule
distinguishable or incoherent alternatives	add probabilities
coherent alternatives leading to the same outcome	add amplitudes, then take a squared modulus

The difference is physical, not semantic. If a detector, environment, or internal degree of freedom records which alternative occurred, the interference term is reduced or removed. If no such record exists and the alternatives remain coherent, the interference term is part of the prediction.

This is the amplitude logic behind the Double-Slit Experiment and broader Interference With Matter.

Scattering Amplitudes

Scattering made amplitudes operational. A beam hits a target, and detectors count particles emerging into directions. Wave mechanics naturally produces outgoing waves, but the experiment records localized counts and rates.

For short-range elastic scattering in a common modern convention, the large-distance state has the form

$$\psi_{\mathbf k}^{(+)}(\mathbf r) \sim e^{i\mathbf k\cdot\mathbf r} + f(\theta,\phi) \frac{e^{ikr}}{r}.$$

The complex function $f(\theta,\phi)$ is the scattering amplitude. The differential cross section is

$$\frac{d\sigma}{d\Omega} = \lvert f(\theta,\phi)\rvert^2$$

in that convention. The formal convention and flux derivation belong to Scattering Amplitude and Differential and Total Cross Sections.

Born’s work showed how a wave calculation could yield probabilities for particle-like scattering events. In weak-potential scattering, the first Born approximation has the schematic structure

$$f_{\mathrm{Born}}(\mathbf k'\leftarrow\mathbf k) \propto \int d^3r\, e^{-i(\mathbf k'-\mathbf k)\cdot\mathbf r} V(\mathbf r).$$

The amplitude is controlled by the Fourier component of the potential at momentum transfer $\mathbf k'-\mathbf k$. The observed cross section involves a squared modulus, but the phase of the amplitude is not disposable. It matters when amplitudes from different channels, paths, partial waves, or identical-particle alternatives are combined.

This is why the word “amplitude” is not just a convenient label. It marks a quantity that can interfere before it becomes an observable probability or rate.

Modern Bridge to Path Integrals and QFT

Dirac’s transformation theory made amplitudes central in a broader language: transition amplitudes such as

$$\langle b\vert a\rangle$$

relate one description, basis, or state to another. Time evolution is also expressed through amplitudes. For position states,

$$K(x_b,t_b;x_a,t_a) = \langle x_b,t_b\vert x_a,t_a\rangle$$

is a propagator amplitude. Propagators compose by summing over intermediate alternatives:

$$K(x_c,t_c;x_a,t_a) = \int dx_b\, K(x_c,t_c;x_b,t_b) K(x_b,t_b;x_a,t_a).$$

Feynman’s path-integral formulation made the same principle vivid: the amplitude is obtained by summing contributions from histories, with phases controlled by the action,

$$\text{amplitude} \sim \sum_{\text{histories}} e^{iS/\hbar}.$$

This is not the historical origin of Born’s 1926 rule, but it is a powerful later expression of amplitude-based reasoning. The formal development starts in Why Path Integrals? and From Propagators to Path Integrals.

The same logic also survives the move to quantum field theory. Scattering calculations produce amplitudes, often denoted by an $S$-matrix element or invariant matrix element. Observable rates and cross sections involve squared moduli together with phase-space factors, flux factors, sums over unobserved quantum numbers, and averages over initial preparations. The bridge pages QFT Bridge: Scattering and QFT Bridge: Born Approximation and Tree Level develop that connection.

The conceptual continuity is simple and deep: quantum theory predicts by assigning amplitudes to processes and probabilities to specified outcomes.

What This Page Does Not Claim

Probability amplitudes do not by themselves settle what the quantum state “really is.” Instrumentalist, realist, hidden-variable, many-worlds, collapse, and relational approaches can agree on the same amplitude calculus while disagreeing about ontology.

Amplitudes also do not mean every alternative in a calculation is a literal path taken by a tiny classical object. Some amplitude decompositions are basis choices, perturbative expansions, histories in a path integral, or diagrammatic terms. The physical question is which amplitudes must be combined coherently for the experimental arrangement.

Finally, measuring only a probability or cross section usually loses phase information. Reconstructing amplitudes often requires interference, additional observables, constraints such as unitarity, or a model.

Common Mistakes

• Treating amplitudes as probabilities.
• Squaring a complex amplitude instead of taking its squared modulus.
• Adding probabilities when the alternatives are coherent.
• Adding amplitudes when the alternatives are physically distinguished or decohered.
• Thinking the phase of an amplitude is irrelevant because probabilities are real.
• Assuming every term in an amplitude expansion is a literal classical trajectory.
• Reading path integrals or Feynman diagrams as probabilities term by term.

Cross-Links

• Probability, Measurement, and Interpretation
• Born Rule as Historical Development
• Complementarity
• Interpreting the Wavefunction
• Double-Slit Experiment
• Interference With Matter
• Dirac’s Transformation Theory
• Probability Amplitudes
• Born Rule
• Transition Probabilities
• Scattering Amplitude
• Why Path Integrals?
• QFT Bridge: Scattering

References

• M. Born, “Zur Quantenmechanik der Stoßvorgänge,” Zeitschrift für Physik 37, 863-867, 1926, DOI: 10.1007/BF01397477.
• P. A. M. Dirac, “The physical interpretation of the quantum dynamics,” Proceedings of the Royal Society A 113, 621-641, 1927, DOI: 10.1098/rspa.1927.0012.
• P. A. M. Dirac, The Principles of Quantum Mechanics, 4th ed., Oxford University Press, 1958.
• P. A. M. Dirac, “The Lagrangian in Quantum Mechanics,” Physikalische Zeitschrift der Sowjetunion 3, 64-72, 1933.
• R. P. Feynman, “Space-time approach to non-relativistic quantum mechanics,” Reviews of Modern Physics 20, 367-387, 1948, DOI: 10.1103/RevModPhys.20.367.
• R. P. Feynman, R. B. Leighton, and M. Sands, The Feynman Lectures on Physics, Volume III, Chapter 1, Addison-Wesley, 1965.
• B. L. van der Waerden, ed., Sources of Quantum Mechanics, Dover, 1968.
• M. Jammer, The Conceptual Development of Quantum Mechanics, 2nd ed., American Institute of Physics, 1989.
• L. E. Ballentine, Quantum Mechanics: A Modern Development, 2nd ed., World Scientific, 2014.

Exercises

1. Two coherent alternatives have amplitudes $A$ and $Ae^{i\phi}$. Derive the probability.

Solution

Add amplitudes first:

$$A_{\mathrm{tot}} = A(1+e^{i\phi}).$$

Then take the squared modulus:

$$\begin{aligned} P &= \lvert A\rvert^2 \lvert 1+e^{i\phi}\rvert^2 \\ &= 2\lvert A\rvert^2(1+\cos\phi). \end{aligned}$$

The result depends on the relative phase $\phi$.

2. Two alternatives are perfectly distinguishable, with individual probabilities $p_1$ and $p_2$. What changes compared with coherent amplitude addition?

Solution

If the alternatives are physically distinguished or incoherent, the interference term is absent. The probability is

$$P = p_1+p_2.$$

One should not write $\lvert A_1+A_2\rvert^2$ unless the alternatives contribute coherently to the same outcome.

3. A scattering experiment measures $d\sigma/d\Omega=\lvert f(\theta,\phi)\rvert^2$. Why does this not generally determine $f(\theta,\phi)$?

Solution

The cross section gives the squared modulus of the amplitude, not its phase:

$$f(\theta,\phi) = \lvert f(\theta,\phi)\rvert e^{i\alpha(\theta,\phi)}.$$

Different phases can give the same $\lvert f\rvert^2$. Phase information may be recovered through interference, polarization observables, unitarity constraints, or a dynamical model, but it is not contained in a single modulus measurement alone.

4. In the propagator composition law, why is there an integral over intermediate positions rather than a sum of probabilities over intermediate positions?

Solution

The intermediate position is not being treated as a recorded measurement outcome. It is an unresolved alternative in the amplitude for propagation from the initial event to the final event. Therefore the contributions are amplitudes:

$$K(x_c,t_c;x_a,t_a) = \int dx_b\, K(x_c,t_c;x_b,t_b) K(x_b,t_b;x_a,t_a).$$

Probabilities would be added only if the intermediate alternatives were physically distinguished or measured in the relevant setup.