Yang–Mills Theory

Yang–Mills theory is the local relativistic theory of a connection for a non-Abelian gauge group. Its curvature contains the gauge potential quadratically, so the same principle that removes redundant polarizations also fixes cubic and quartic gauge-boson interactions. This chapter follows that fact from the classical action and Gauss constraint through ghosts, color vertices, asymptotic freedom, the background-field method, and the BRST statement of perturbative consistency.

Enter Yang–Mills theory

Observable-readiness check. First identify the object whose value you want. A closed Wilson loop, a color-singlet local composite, or an amplitude between BRST-physical states can be an observable; $A_\mu^a$, a colored Green function, and a gauge-fixed effective action are intermediate objects. If that distinction is not yet operational, repair it at Yang–Mills equations, constraints, and observables before interpreting a calculation. If the target is a covariant-gauge amplitude but ghosts and the physical-state condition are not yet available, repair that gap at Gauge-fixed Yang–Mills action and ghost sector.

Choose a route by the object that must be constructed or checked:

Question	Start here	Result carried forward
Why do gauge bosons interact with one another?	The Yang–Mills action and gauge self-interaction	Curvature, action, equations, Bianchi identity, and the cubic and quartic terms
Which data are constrained, and which quantities are observable?	Yang–Mills equations, constraints, and observables	Gauss law, first-class constraint algebra, two local polarizations, and Wilson observables
Why are ghosts required in a covariant path integral?	Gauge-fixed Yang–Mills action and ghost sector	Faddeev–Popov operator, propagators, BRST differential, and the physical-state handoff
How are the perturbative rules normalized and checked?	Yang–Mills color algebra and perturbative vertices	Group invariants, three- and four-gauge vertices, the ghost vertex, and Ward checks
When is the theory weakly coupled at short distance?	Non-Abelian screening and asymptotic freedom	The one-loop coefficient with fermion and scalar matter, its sign, and the generated scale
How can one keep gauge covariance visible in loop calculations?	Background-field Yang–Mills effective action	Background-covariant operators, the background Ward identity, and a short beta-function route
What exactly do renormalizability and unitarity mean here?	Consistency of perturbative Yang–Mills theory	Slavnov–Taylor-compatible counterterms and unitarity on BRST cohomology

The table order is a suggested route, not a second prerequisite system. Each leaf’s Required background note is the hard gate for that page; its Helpful background note is optional. Readers who can already derive the covariant gauge and ghost action may enter at color vertices. The background-field and perturbative-consistency pages still require the BRST and counterterm capabilities linked on those pages, even if the intervening applications are skipped.

One structure, several descriptions

With Hermitian generators and the site-wide convention

$$D_\mu=\partial_\mu-igA_\mu, \qquad [D_\mu,D_\nu]=-igF_{\mu\nu},$$

the component curvature is

$$F^a_{\mu\nu} =\partial_\mu A^a_\nu-\partial_\nu A^a_\mu +g f^{abc}A^b_\mu A^c_\nu.$$

That last term is the source of the chapter’s entire perturbative spine. Squaring $F$ produces three- and four-gauge-boson vertices. The same structure constants enter the Gauss constraint and its closed algebra. Covariant gauge fixing differentiates the gauge transformation and therefore produces an interacting adjoint ghost. At one loop, gauge and ghost fluctuations combine into the $+11C_A/3$ contribution to the coefficient $b_0$, opposite in sign to matter screening. BRST symmetry then relates the counterterms of all these apparently different vertices.

This is not merely a chain of analogies: every arrow is checked by an identity. The Jacobi identity gives the Bianchi identity and BRST nilpotency; contraction of the three-gauge vertex gives a difference of inverse kinetic operators; background gauge invariance gives $Z_g Z_{\bar A}^{1/2}=1$; and the Slavnov functional identity restricts the counterterm space. The original nonlinear gauge-field construction appears in Yang and Mills 1954, pp. 191–195, while a unified modern derivation is given in Schwartz 2014, Chapters 25–26, pp. 481–533.

What the chapter establishes

For a declared compact group, representations, generator normalization, and gauge-fixing convention, the chapter establishes the perturbative Yang–Mills model around a specified vacuum:

• the classical equations, constraints, local degree count, and basic gauge-invariant observables;
• covariant propagators and interaction vertices, including the ghost sector;
• the universal one-loop sign of the beta function and its matter-content boundary;
• the background-field identity that reduces coupling renormalization to a background two-point calculation; and
• the conditions under which the renormalized $S$-matrix descends to a unitary operator on the perturbative physical state space.

Several conclusions lie outside that statement. A Lie algebra and local action do not by themselves choose the global form of the gauge group, the spectrum of genuine line operators, or a topological $\theta$ sector. A negative one-loop beta function does not prove confinement, a mass gap, or even a particular infrared phase. Perturbative BRST arguments do not construct the theory nonperturbatively or remove global gauge-fixing obstructions. Those boundaries are part of the scientific result, not qualifications to be discarded.

A compact consistency test

A calculation is ready to leave this chapter when all four questions have definite, checkable answers:

1. Conventions: $[T^a,T^b]$, $\operatorname{tr}(T^aT^b)$, $D_\mu$, and the Fourier sign predict the same cubic vertex. If not, return to the action convention sheet.
2. Identities: contracting the three-gauge vertex gives a difference of inverse transverse kernels, and the color tensors satisfy Jacobi and Casimir identities. If not, repair the rules at color algebra and perturbative vertices.
3. Observable status: the reported quantity is either gauge invariant or explicitly labeled as an intermediate Green function, and a physical amplitude is independent of $\xi$. If not, return to generators, charges, and observables and the BRST physical state space.
4. Claim ceiling: the conclusion states its perturbative order and does not infer an infrared phase, nonperturbative existence, or anomaly cancellation from a UV loop calculation. If not, use the one-loop interpretation boundary and the consistency boundary.

Failure of any one check usually points to a missing ghost contribution, a reversed coupling sign, an inconsistent generator normalization, or an observable that is still gauge dependent.

Continue by purpose

• To specialize the gauge sector to quarks, partons, factorization, and collider observables, continue to Perturbative QCD and Partons.
• To study global form, line operators, screening, and phase diagnostics, continue to Gauge Dynamics, Charges, and Phases.
• To develop general amplitude construction beyond this chapter’s vertex checks, continue to Scattering Theory.
• To treat BRST, BV, Gribov copies, and anomalies as primary subjects, continue to Gauge Fixing, BRST, BV, and Anomalies.
• To place the beta function and counterterms in their general RG and EFT setting, continue to Renormalization and Effective Field Theory.

References

• Matthew D. Schwartz, Quantum Field Theory and the Standard Model, Cambridge University Press (2014), Chapters 25–26 and § 34.3, DOI.
• C. N. Yang and R. L. Mills, “Conservation of Isotopic Spin and Isotopic Gauge Invariance,” Physical Review 96 (1954), 191–195, DOI.