Chapter 31 Key Takeaways

The Big Ideas

1. The Path Integral Is an Exact Reformulation of Quantum Mechanics

The propagator $K(x_f, t; x_i, 0) = \int \mathcal{D}[x]\, e^{iS[x]/\hbar}$ is derived from the time-evolution operator $\hat{U}(t) = e^{-i\hat{H}t/\hbar}$ by inserting completeness relations. No approximations are made. Every prediction of the Schrödinger equation can be recovered from the path integral, and vice versa.

What to remember: The Schrödinger equation and the path integral are two languages for the same physics. Fluency in both is essential for modern theoretical physics.

2. Every Path Contributes — The Classical Path Is Not Special (Until It Is)

In the path integral, every continuous path from $x_i$ to $x_f$ contributes with equal magnitude $|e^{iS/\hbar}| = 1$. No path is excluded, no matter how wild. The classical path has no privileged status in the formulation.

What makes the classical path special is the stationary phase condition: when $S_{\text{cl}} \gg \hbar$, paths near the classical trajectory have nearly identical phases and add constructively, while distant paths cancel through destructive interference. Classical mechanics is an interference effect.

What to remember: Quantum mechanics sums all paths democratically. Classical mechanics is what survives after the cancellations.

3. The Classical Limit Is Immediate and Transparent

The stationary phase approximation $\delta S = 0$ directly gives the Euler-Lagrange equations — Newton's second law. The principle of least action, the foundation of classical mechanics, emerges as the constructive interference condition of the path integral. No Ehrenfest theorem, no narrow-wave-packet assumption, no additional arguments needed.

What to remember: Hamilton's principle of stationary action is not an independent postulate of classical mechanics. It is a theorem of quantum mechanics, derivable from the path integral in the limit $S \gg \hbar$.

4. Gaussian Path Integrals Can Be Evaluated Exactly

When the action is quadratic in $x$ and $\dot{x}$ — as for the free particle, the QHO, and particles in uniform fields — the path integral is Gaussian and can be computed exactly. The result factorizes as $K = A(x_f, x_i, t)\, e^{iS_{\text{cl}}/\hbar}$, where $A$ is determined by the second variation (fluctuation determinant).

What to remember: Free particle and QHO path integrals are the two workhorses. All perturbative calculations are built on top of them.

5. The QHO Propagator Encodes the Entire Energy Spectrum

The QHO propagator $K = \sqrt{m\omega/(2\pi i\hbar\sin\omega t)}\,\exp[\ldots]$ contains the energy eigenvalues $E_n = \hbar\omega(n+1/2)$ in its analytic structure — specifically, in the poles of $1/\sin(\omega t)$. The path integral derives the QHO spectrum without solving a differential equation or using ladder operators.

What to remember: The propagator is a generating function for the energy spectrum. Taking its trace (and Wick-rotating) yields the partition function.

6. The Wick Rotation Connects Quantum Mechanics to Statistical Mechanics

Replacing $t \to -i\tau$ transforms the oscillatory path integral $\int \mathcal{D}[x]\, e^{iS/\hbar}$ into the convergent Euclidean path integral $\int \mathcal{D}[x]\, e^{-S_E/\hbar}$. The quantum partition function $Z = \text{Tr}(e^{-\beta\hat{H}})$ is a Euclidean path integral over periodic paths of period $\beta\hbar$.

What to remember: Quantum mechanics in $d$ dimensions = classical statistical mechanics in $d+1$ dimensions. This is the basis of lattice gauge theory, path integral Monte Carlo, and the instanton calculus for tunneling.

7. Feynman Diagrams Emerge from Perturbative Expansion

Expanding the non-Gaussian part of the path integral in a Taylor series and applying Wick's theorem yields a sum of terms, each representable as a diagram. Lines are propagators, vertices are interaction couplings, and the Feynman rules translate pictures into integrals.

What to remember: Feynman diagrams are not mnemonics — they are terms in the path integral's perturbation series. This is why they extend naturally from quantum mechanics to quantum field theory.

Key Equations to Internalize

  1. Path integral: $K(x_f, t; x_i, 0) = \int \mathcal{D}[x]\, e^{iS[x]/\hbar}$ — The central formula. Memorize it.

  2. Free-particle propagator: $K_{\text{free}} = \sqrt{\frac{m}{2\pi i\hbar t}}\exp\left[\frac{im(x_f - x_i)^2}{2\hbar t}\right]$ — The simplest exact path integral.

  3. QHO propagator: $K_{\text{QHO}} = \sqrt{\frac{m\omega}{2\pi i\hbar\sin\omega t}}\exp\left[\frac{im\omega}{2\hbar\sin\omega t}((x_i^2 + x_f^2)\cos\omega t - 2x_i x_f)\right]$

  4. QHO partition function: $Z = 1/[2\sinh(\beta\hbar\omega/2)]$ — From the Wick-rotated trace.

  5. Classical limit: $\delta S[x_{\text{cl}}] = 0$ — Stationary phase selects the classical path.

Common Mistakes to Avoid

  • Thinking the path integral is an approximation. It is exact. The derivation from the time-evolution operator uses only the completeness relation and the Trotter formula (which becomes exact in the $N \to \infty$ limit).

  • Confusing "all paths" with "all classical paths." The path integral sums over all continuous paths, not just solutions of the equations of motion. The classical path is one among uncountably many. Most paths in the integral are nowhere-differentiable.

  • Forgetting the normalization prefactor. The path integral measure $\mathcal{D}[x]$ includes the factor $(m/(2\pi i\hbar\epsilon))^{N/2}$. Without it, the propagator has the wrong dimensions and wrong normalization.

  • Misinterpreting the Wick rotation. The replacement $t \to -i\tau$ is a mathematical technique for evaluating integrals, not a physical claim that "time is imaginary." Real-time and imaginary-time path integrals describe different physical quantities (dynamics vs. thermodynamics).

  • Thinking Feynman diagrams are merely pictures. Each diagram is a precise mathematical expression — an integral over vertex positions (or momenta) with specific propagators and coupling constants. The rules for translating diagrams to integrals follow rigorously from Wick's theorem.

Connections Forward

  • Chapter 32 (Adiabatic Theorem and Berry Phase): The Berry phase can be derived from the path integral in the adiabatic limit. The geometric phase is the holonomy of a connection on the space of Hamiltonians — naturally formulated using the path integral.

  • Chapter 34 (Second Quantization): The field-theoretic path integral integrates over field configurations $\phi(\mathbf{x}, t)$ rather than particle trajectories $x(t)$. The same structure — action, Gaussian integrals, Wick's theorem — carries over directly.

  • Chapter 36 (Topological Phases): Topological contributions to the path integral (winding numbers, Chern-Simons terms) are naturally formulated in the path integral language. The Aharonov-Bohm effect is the simplest example.

  • Chapter 37 (From QM to QFT): The path integral is the natural bridge. Feynman rules, renormalization, and the Standard Model are all built on the perturbative expansion of the field-theory path integral developed in this chapter.

Connections Backward

  • Chapter 4 (QHO): The anchor example. The QHO energy spectrum $E_n = \hbar\omega(n+1/2)$ is now re-derived from the poles of the propagator — a completely independent method from the ladder operators of Chapter 4.

  • Chapter 7 (Time Evolution): The time-evolution operator $\hat{U}(t) = e^{-i\hat{H}t/\hbar}$ is the starting point of the path integral derivation. The propagator $K = \langle x_f|\hat{U}|x_i\rangle$ is the position-space matrix element we have been computing since Chapter 7.

  • Chapter 8 (Dirac Notation): The completeness relation $\hat{I} = \int dx\,|x\rangle\langle x|$ — the central tool of Chapter 8 — is the key mathematical step in the path integral derivation. Without it, the path integral cannot be constructed.