Why Path Integrals?

Path integrals express quantum transition amplitudes as sums over histories weighted by the action phase:

$$e^{iS/\hbar}.$$

They are powerful because they connect quantum dynamics to the classical action while preserving interference.

From Wavefunctions to Amplitudes

The propagator

$$K(x_f,t_f;x_i,t_i) = \langle x_f\rvert U(t_f,t_i)\lvert x_i\rangle$$

is an amplitude density in the position basis. It evolves a wavefunction by integration over its initial position; exact position kets are not normalized physical preparations. The path integral rewrites this kernel as a formal sum over paths connecting the endpoints.

Sum Over Alternatives

Quantum amplitudes add over indistinguishable alternatives. In a multi-slit experiment, amplitudes from different slits add. In the path-integral viewpoint, the alternatives are refined until they become entire paths.

Formally,

$$K(x_f,t_f;x_i,t_i) = \int_{x(t_i)=x_i}^{x(t_f)=x_f} \mathcal D x(t)\, e^{iS[x]/\hbar}.$$

This is not an ordinary probability sum. The phases interfere.

Classical Action as Phase

The action is

$$S[x]=\int_{t_i}^{t_f}L(x,\dot x,t)\,dt.$$

When action differences across a family of paths are large compared with $\hbar$, their phases vary rapidly. An additive constant in the action has no such effect. Contributions away from stationary-action paths tend to cancel, while neighborhoods of classical paths add coherently. This is the stationary-phase route from quantum amplitudes to classical mechanics; see Semiclassical Limit Overview for the Core-level bridge and Semiclassical Limit for the Toolkit-level mechanism.

The same action can appear in the opposite direction when path integrals are used as a quantization strategy. Quantization vs Classical Limit keeps those two uses separate.

Why the Formulation Is Powerful

Path integrals are useful because they:

• make the action central;
• expose the classical limit through stationary phase;
• treat many degrees of freedom in a way that generalizes naturally to fields;
• organize perturbation theory with sources and correlation functions;
• connect quantum mechanics to statistical mechanics through imaginary time;
• make some symmetry and topology effects easier to see.

Why the Formulation Is Subtle

Real-time path integrals are oscillatory, not probability measures. The notation $\mathcal D x(t)$ hides limiting procedures, normalization factors, operator-ordering choices, and convergence questions. A path-integral expression should be checked against operator definitions and boundary conditions.

When sources are introduced, Functional Derivatives provide the compact notation for generating coordinate or field insertions. Their physical meaning depends on the operator ordering and boundary conditions of the path-integral construction.

Common Mistakes

• Saying particles literally travel along every path as a classical statement.
• Treating $e^{iS/\hbar}$ as a probability weight.
• Ignoring normalization and measure factors.
• Assuming path integrals are automatically rigorous measures.
• Forgetting that operator ordering can affect the path-integral form.

References

• R. P. Feynman and A. R. Hibbs, Quantum Mechanics and Path Integrals, McGraw-Hill, 1965.
• L. S. Schulman, Techniques and Applications of Path Integration, Wiley, 1981.
• R. Shankar, Principles of Quantum Mechanics, 2nd ed., Springer, 1994.
• J. Zinn-Justin, Path Integrals in Quantum Mechanics, Oxford University Press, 2005.

Exercises

Two alternatives contribute $a e^{iS_1/\hbar}$ and $a e^{iS_2/\hbar}$ with the same real coefficient $a$. Compare their summed intensity when $S_2-S_1=\pi\hbar$ and when $S_2-S_1=2\pi\hbar$. Does adding a common action constant affect either answer?

Solution

The summed amplitude is $a e^{iS_1/\hbar}(1+e^{i\Delta S/\hbar})$. For $\Delta S=\pi\hbar$ it vanishes; for $\Delta S=2\pi\hbar$ its squared magnitude is $4a^2$. A common shift multiplies both contributions by the same unit phase and leaves the intensity unchanged. This is why action differences, rather than the absolute size of $S$, govern interference.