58
Part Two · Quantum fields and particles

Classical field theory

Picture a mattress: a grid of springs, one at every point, each tied to its neighbours. Press one point and a ripple spreads. Now make the springs smaller and closer together, without limit. What is left is a field, a number at every point of space that wobbles and passes its wobble on. Every quantum field in the Standard Model starts as one of these, written down with a single line, its Lagrangian density.

On the plate: a chain of masses joined by springs. Increase the number of masses and the chain becomes a continuous string.

From beads to a field

Take NN beads of mass MM spaced aa apart, joined by springs of stiffness κ\kappa. Each bead obeys Newton’s law with forces from its two neighbours. Let a→0a \to 0 with the mass per length μ=M/a\mu = M/a and the tension T=κaT = \kappa a fixed. The bead displacements qi(t)q_i(t) become a field ϕ(x,t)\phi(x, t), and the difference between neighbours becomes a derivative. The equation of motion becomes the wave equation of Chapter 5, with speed T/μ\sqrt{T/\mu}.

Use 16 beads. With few beads the pulse breaks into jitter; with many it travels cleanly as a continuous wave.

Lagrangian density

For the beads, the Lagrangian of Chapter 12 is a sum over beads of kinetic minus potential energy. In the limit the sum becomes an integral, L=∫L dxL = \int\mathcal L\,dx, and the action is S=∫L d4xS = \int\mathcal L\,d^4x. For a relativistic scalar field the simplest choice is

L=12 ∂μϕ ∂μϕ−12m2ϕ2\mathcal L = \tfrac12\,\partial_\mu\phi\,\partial^\mu\phi - \tfrac12 m^2\phi^2

Demanding δS=0\delta S = 0 for all variations gives the field version of the Euler–Lagrange equation, ∂μ∂L∂(∂μϕ)=∂L∂ϕ\partial_\mu\frac{\partial\mathcal L}{\partial(\partial_\mu\phi)} = \frac{\partial\mathcal L}{\partial\phi}, which here is Klein–Gordon, Chapter 56. The mass term is a spring tying each bead to the ground. Established

What the mass term does

A wave of wavenumber kk on the bead chain has frequency ω2=ω02+4κMsin⁡2(ka/2)\omega^2 = \omega_0^2 + \frac{4\kappa}{M}\sin^2(ka/2), where ω0\omega_0 comes from the springs to the ground. Set ω0=\omega_0 = 1.0. For small kk the curve becomes ω2=ω02+c2k2\omega^2 = \omega_0^2 + c^2k^2: exactly E2=m2+p2E^2 = m^2 + p^2 with E=ℏωE = \hbar\omega. The grid spacing shows only at wavenumbers near π/a\pi/a. Whether real space has a smallest spacing is unknown; Part Five returns to it.

Three Lagrangians for the whole Standard Model

Every field you will meet is built from three kinds of term.

Lscalar=12(∂ϕ)2−12m2ϕ2LDirac=ψˉ (iγμ∂μ−m) ψLMaxwell=−14FμνFμν−jμAμ\begin{gathered}\mathcal L_{\text{scalar}} = \tfrac12(\partial\phi)^2 - \tfrac12m^2\phi^2 \\ \mathcal L_{\text{Dirac}} = \bar\psi\,(i\gamma^\mu\partial_\mu - m)\,\psi \\ \mathcal L_{\text{Maxwell}} = -\tfrac14F_{\mu\nu}F^{\mu\nu} - j^\mu A_\mu\end{gathered}

Vary each and you recover Klein–Gordon, Dirac (Chapter 57) and Maxwell’s four equations (Chapter 15), with Fμν=∂μAν−∂νAμF_{\mu\nu} = \partial_\mu A_\nu - \partial_\nu A_\mu and ψˉ=ψ†γ0\bar\psi = \psi^\dagger\gamma^0. The plate shows a small change to a field configuration and the change in action it produces; on a solution, every small change leaves the action unchanged to first order. Interactions are extra products of fields, such as ψˉγμψAμ\bar\psi\gamma^\mu\psi A_\mu, which couples electrons to light (Chapter 67).

Why Lagrangians

Writing equations of motion directly would also work. Lagrangians win for three reasons. Symmetry is visible: if L\mathcal L is unchanged by a transformation, so is the physics, and Noether’s theorem (next chapter) turns each symmetry into a conservation law. Relativity is automatic: if L\mathcal L is a Lorentz scalar, the equations are covariant. And quantization starts from it, through canonical momenta (Chapter 60) or the path integral of Chapter 53, where eiSe^{iS} weighs each field configuration.

What you now have

Fields as limits of coupled oscillators, the Lagrangian density and the field Euler–Lagrange equation, the mass term as a spring to the ground, the scalar, Dirac and Maxwell Lagrangians, and why physics is written this way.

· · ·

Three questions to carry with you.

What does the mass term correspond to on the bead chain?tap
A spring tying each bead to a fixed ground, giving a minimum frequency even at k = 0.
What equation does ½(∂φ)² − ½m²φ² produce?tap
The Klein–Gordon equation, (∂² + m²)φ = 0.
Where does the grid spacing show up in the dispersion?tap
Near k ≈ π/a, where the sine bends away from the straight relativistic curve.

And one to work out. A string has tension 100 N and mass per length 0.01 kg/m. What is its wave speed T/μ\sqrt{T/\mu}, in m/s?

v =m/s

Sources: H. Goldstein, C. Poole and J. Safko, Classical Mechanics, ch. 13. D. Tong, Lectures on Quantum Field Theory, ch. 1. M. D. Schwartz, Quantum Field Theory and the Standard Model, ch. 3. Checked by code: bead-chain dispersion ω² = ω₀² + (4κ/M) sin²(ka/2) against direct normal-mode diagonalization; small-k limit ω² = ω₀² + c²k².

Next · Chapter 59
Noether's theorem

Energy is conserved because the laws of physics are the same today as tomorrow.

Turn the page