Gauge Fixing and Faddeev-Popov Ghosts in Non-Abelian Quantization

The Problem: An Integral Over Infinite Redundancy

The starting point for path integral quantization is the partition function, $Z$, from which we can calculate correlation functions. For a gauge theory, we might naively write:

where the integral is over all possible configurations of the gauge field $A_{\mu }$. The problem lies in the gauge invariance of the action, $S_{\text{YM}}[A^g]=S_{\text{YM}}[A]$, where $A^g$ is a gauge-transformed version of $A$.

For any given field configuration $A$, there exists an infinite set of other configurations—the gauge orbit of $A$—that are physically equivalent and give the exact same value for the action. The path integral, as written above, sums over all these redundant configurations. It's like integrating a constant function over an infinite volume, which leads to a divergent result. We are massively overcounting.

To make progress, we need a method to isolate a single, unique representative from each gauge orbit and integrate only over these physically distinct configurations. This procedure is known as gauge fixing.

The video below provides a concise explanation of this overcounting problem.

The Faddeev-Popov Method

The solution, developed by Ludvig Faddeev and Victor Popov, is a masterful trick. It allows us to restrict the path integral to a "slice" through the space of field configurations that (ideally) picks out one representative per gauge orbit, while simultaneously introducing a correction factor to ensure the final result is independent of the specific slice chosen.

The method involves three conceptual steps:

1. Choose a Gauge-Fixing Condition: We define a condition, $G[A]=0$, that specifies the slice. A common choice, analogous to the one in QED, is the Lorenz gauge condition:

   $$G^a[A]={\partial }^{\mu }{A}_{\mu }^a(x)=0$$

   This condition must be imposed for each component $a$ of the gauge field and at every spacetime point $x$.

2. Insert a Clever "1": We insert an identity into the path integral. This identity is constructed using the gauge-fixing condition and a special determinant, now known as the Faddeev-Popov determinant, ${\Delta }_{FP}[A]$. The identity is:

   $$1={\Delta }_{FP}[A]\int \mathcal{D}g\delta (G[A^g])$$

   Here, the integral is over all possible gauge transformations $g$, and the delta functional $\delta (G[A^g])$ enforces the gauge condition on the transformed field $A^g$. The determinant ${\Delta }_{FP}[A]$ is precisely the Jacobian factor required to make this identity hold and, crucially, to make the path integral independent of the choice of $G$.

3. Manipulate the Path Integral: By inserting this "1" into the path integral and using the gauge invariance of the action and the integration measure, one can factor out the infinite volume of the gauge group, leaving a well-defined integral over a gauge-fixed action.

The following text provides an excellent, pedagogically sound walkthrough of this procedure, starting with a finite-dimensional analogy that builds intuition for the infinite-dimensional case of field theory. Your background in mathematics will make the logic of this geometric approach particularly clear.

Faddeev-Popov Ghosts: An Unphysical Necessity

As you just read, the Faddeev-Popov procedure confronts us with the determinant of a field-dependent differential operator, ${\Delta }_{FP}[A]=\det ({\partial }^{\mu }{D}_{\mu })$, sitting inside our path integral. This non-local, complicated term would make perturbative calculations impossible.

The second brilliant insight of Faddeev and Popov was to represent this determinant itself as a path integral, using the properties of Gaussian integrals for Grassmann numbers (anti-commuting variables).

By applying this identity, the troublesome determinant is replaced by an exponential of a new, local action term involving two new fields, $c(x)$ and $\bar{c}(x)$. These are the Faddeev-Popov ghosts.

Let's be very clear about their properties:

• Lorentz Nature: They are complex scalar fields.
• Statistics: They obey Fermi-Dirac statistics (they are anti-commuting Grassmann fields).

This combination directly violates the spin-statistics theorem, which demands that integer-spin fields (like scalars) be bosons and half-integer spin fields be fermions. This violation is the definitive sign that ghosts are not, and can never be, physical particles. They are purely a mathematical tool required for a consistent quantization of the theory. They appear only as internal lines in Feynman diagrams, never as external, observable states.

The Gauge-Fixed Lagrangian and Feynman Rules

To make the path integral suitable for perturbation theory, one final trick is employed. The rigid delta functional $\delta ({\partial }^{\mu }{A}_{\mu }^a)$ is "softened" by averaging over a family of gauge conditions. This results in adding a gauge-fixing term to the Lagrangian:

Here, $\xi$ is an arbitrary real number called the gauge-fixing parameter. This family of gauges is known as $R_{\xi }$ gauges. Physical quantities must be independent of the choice of $\xi$.

Putting everything together, the effective Lagrangian for the quantum theory, which we use to derive Feynman rules, is:

Expanding the ghost term, ${\mathcal{L}}_{\text{ghost}}=({\partial }_{\mu }{\bar{c}}^a)({\partial }^{\mu }{c}^a)+g{f}^{abc}({\partial }_{\mu }{\bar{c}}^a)A_b^{\mu }{c}^c$, reveals a kinetic term for the ghosts and, crucially, a new ghost-gluon interaction vertex.

This effective Lagrangian gives rise to a set of Feynman rules:

• A well-defined gluon propagator (invertible thanks to the gauge-fixing term).
• A ghost propagator.
• The 3-gluon and 4-gluon self-interaction vertices from ${\mathcal{L}}_{\text{YM}}$.
• A new antighost-gluon-ghost vertex.

The unphysical longitudinal polarizations of the gluon and the unphysical ghost particles both contribute to loop diagrams. Miraculously, their contributions are arranged to exactly cancel each other out in the calculation of any physical observable, such as an S-matrix element between physical external states. This cancellation ensures the unitarity of the theory. The deep symmetry responsible for this cancellation is known as BRST symmetry.

A Note on the Gribov Ambiguity

It is important to mention a subtlety that was discovered after the original work of Faddeev and Popov. The assumption that a gauge-fixing surface like ${\partial }^{\mu }{A}_{\mu }^a=0$ intersects each gauge orbit exactly once turns out to be false for non-abelian theories. It's possible for an orbit to intersect the surface multiple times. These additional intersections are called Gribov copies.

This "Gribov ambiguity" implies that the Faddeev-Popov procedure is not non-perturbatively perfect. However, for the purposes of perturbation theory, where we study small fluctuations around the vacuum $A_{\mu }=0$, we can effectively ignore this issue. It remains a topic of active research in the non-perturbative understanding of Yang-Mills theories.

Conclusion

In this lesson, we have confronted the primary obstacle to quantizing Yang-Mills theory and explored the elegant solution provided by the Faddeev-Popov method.

Key Takeaways:

• The gauge invariance of the Yang-Mills action leads to an infinite overcounting in the path integral, making it ill-defined.
• The Faddeev-Popov procedure fixes this by restricting the integral to a gauge slice and introducing a compensating determinant factor, ${\Delta }_{FP}$.
• The Faddeev-Popov determinant is non-local and field-dependent, but it can be expressed as a path integral over new, unphysical fields called Faddeev-Popov ghosts.
• Ghosts are anti-commuting scalar fields that violate the spin-statistics theorem, confirming they are a mathematical artifact, not physical particles.
• The final, gauge-fixed Lagrangian includes the original Yang-Mills term, a gauge-fixing term (e.g., in $R_{\xi }$ gauge), and a ghost action term that includes a ghost-gluon interaction.
• The unphysical contributions from ghosts and longitudinal gluons cancel in physical observables, ensuring unitarity. This cancellation is enforced by an underlying BRST symmetry.
• Perturbative quantization is well-defined, but the procedure has non-perturbative subtleties related to the Gribov ambiguity.

We now have all the necessary components for a perturbative quantization of non-abelian gauge theories. The price for handling the gauge symmetry was the introduction of a more complex Lagrangian with unphysical ghost fields, whose sole purpose is to enforce consistency.

Preview of the Next Lesson:

With the full machinery of Yang-Mills theory in hand, we are ready to see how it fits into the larger picture of fundamental interactions. In the next lesson, we will assemble the pieces of the Standard Model of Particle Physics, identifying its gauge group structure, SU(3)xSU(2)xU(1), and the roles played by the different gauge theories within it. We will then turn to another crucial set of symmetries: the discrete symmetries of parity (P), charge conjugation (C), and time reversal (T).