Soft Symmetry Improvement in 2PIEA

• Soft Symmetry Improvement is a method in quantum field theory that employs a quadratic penalty to control, rather than strictly enforce, Ward identity constraints in 2PI effective actions.
• It introduces a stiffness parameter that enables interpolation between unimproved formulations with symmetry violations and strict symmetry improvement, as demonstrated in the O(N) scalar model using Hartree–Fock approximations.
• SSI overcomes limitations of hard enforcement methods by avoiding pathological solution behavior and offering controlled symmetry violations, with significant implications for phase transition studies and infrared sensitivity.

Soft Symmetry Improvement (SSI) is a methodology for two-particle-irreducible effective actions (2PIEAs) in quantum field theory, designed to address the explicit violation of Ward identities (WIs) that arises when truncating the 2PIEA of theories with spontaneously broken symmetries. SSI penalizes but does not strictly eliminate WI violations, introducing a stiffness parameter that allows interpolation between the unimproved and strictly symmetry-improved (SI) formulations. This approach aims to overcome the practical and conceptual limitations of hard symmetry improvement, particularly the non-existence of solutions in certain regimes, by enabling a controlled, least-squares relaxation of symmetry constraints (Brown et al., 2016).

1. Formal Construction of the Soft Symmetry Improved 2PIEA

Let $\phi_a(x)$ denote the fields of the theory, $\varphi_a = \langle \phi_a \rangle$ their expectation values, and $\Delta_{ab}$ the two-point function. The standard unimproved 2PIEA is

$$\Gamma[\varphi, \Delta] = S[\varphi] + \frac{1}{2}\mathrm{Tr} \ln \Delta^{-1} + \frac{1}{2}\mathrm{Tr}\left[(\Delta_0^{-1}[\varphi]\,\Delta - 1)\right] + \Gamma_2[\varphi, \Delta],$$

where $S[\varphi]$ is the classical action and $\Gamma_2$ sums 2PI vacuum graphs. Theories with global symmetry generated by $T_{ab}^A$ satisfy the WI

$$0 = \mathcal{W}_a^A[\varphi,\Delta] \equiv (\Delta^{-1})_{ab}T_{bc}^A\varphi_c.$$

Strict SI attempts to impose $\mathcal{W}=0$ via Lagrange multipliers, often causing pathological solution behavior. SSI introduces instead a quadratic penalty for WI violation: $$\Gamma_\xi^{\mathrm{SSI}}[\varphi, \Delta] = \Gamma[\varphi, \Delta] + \frac{1}{2}\xi^{-1}\|\mathcal{W}[\varphi, \Delta]\|^2,$$ with $\varphi_a = \langle \phi_a \rangle$0 controlling the penalty. Typically,

$\varphi_a = \langle \phi_a \rangle$1

though a more general kernel $\varphi_a = \langle \phi_a \rangle$2 can be introduced, usually chosen to be local and diagonal. The equations of motion are derived by extremizing $\varphi_a = \langle \phi_a \rangle$3 with respect to $\varphi_a = \langle \phi_a \rangle$4 and $\varphi_a = \langle \phi_a \rangle$5, leading to modified gap equations where the penalty term yields nontrivial contributions proportional to $\varphi_a = \langle \phi_a \rangle$6.

2. Interpolation between Unimproved and Strict Symmetry Improvement

SSI provides a continuous framework interpolating between the unimproved 2PIEA and strict SI. In the formal limit $\varphi_a = \langle \phi_a \rangle$7, the penalty term vanishes, so $\varphi_a = \langle \phi_a \rangle$8 reduces to the unimproved 2PIEA, which generally violates WIs. In the opposite limit $\varphi_a = \langle \phi_a \rangle$9, the penalty fully dominates, strictly enforcing $\Delta_{ab}$0 as in SI.

$\Delta_{ab}$1

For finite $\Delta_{ab}$2, the penalty term allows controlled violations of $\Delta_{ab}$3, offering practical flexibility for finding nonpathological solutions with physically reasonable, if softly broken, symmetry properties.

3. Application to the $\Delta_{ab}$4 Scalar Model and Ward Identities

The SSI construction is concretely realized in the $\Delta_{ab}$5 scalar model in $\Delta_{ab}$6 Euclidean dimensions: $\Delta_{ab}$7 Spontaneous symmetry breaking with vacuum expectation $\Delta_{ab}$8 yields a single Higgs ($\Delta_{ab}$9) and $$\Gamma[\varphi, \Delta] = S[\varphi] + \frac{1}{2}\mathrm{Tr} \ln \Delta^{-1} + \frac{1}{2}\mathrm{Tr}\left[(\Delta_0^{-1}[\varphi]\,\Delta - 1)\right] + \Gamma_2[\varphi, \Delta],$$0 Goldstone fields ($$\Gamma[\varphi, \Delta] = S[\varphi] + \frac{1}{2}\mathrm{Tr} \ln \Delta^{-1} + \frac{1}{2}\mathrm{Tr}\left[(\Delta_0^{-1}[\varphi]\,\Delta - 1)\right] + \Gamma_2[\varphi, \Delta],$$1, $$\Gamma[\varphi, \Delta] = S[\varphi] + \frac{1}{2}\mathrm{Tr} \ln \Delta^{-1} + \frac{1}{2}\mathrm{Tr}\left[(\Delta_0^{-1}[\varphi]\,\Delta - 1)\right] + \Gamma_2[\varphi, \Delta],$$2). The fundamental generators of $$\Gamma[\varphi, \Delta] = S[\varphi] + \frac{1}{2}\mathrm{Tr} \ln \Delta^{-1} + \frac{1}{2}\mathrm{Tr}\left[(\Delta_0^{-1}[\varphi]\,\Delta - 1)\right] + \Gamma_2[\varphi, \Delta],$$3 are $$\Gamma[\varphi, \Delta] = S[\varphi] + \frac{1}{2}\mathrm{Tr} \ln \Delta^{-1} + \frac{1}{2}\mathrm{Tr}\left[(\Delta_0^{-1}[\varphi]\,\Delta - 1)\right] + \Gamma_2[\varphi, \Delta],$$4. The relevant WI reads

$$\Gamma[\varphi, \Delta] = S[\varphi] + \frac{1}{2}\mathrm{Tr} \ln \Delta^{-1} + \frac{1}{2}\mathrm{Tr}\left[(\Delta_0^{-1}[\varphi]\,\Delta - 1)\right] + \Gamma_2[\varphi, \Delta],$$5

for the zero mode of the Goldstone propagator $$\Gamma[\varphi, \Delta] = S[\varphi] + \frac{1}{2}\mathrm{Tr} \ln \Delta^{-1} + \frac{1}{2}\mathrm{Tr}\left[(\Delta_0^{-1}[\varphi]\,\Delta - 1)\right] + \Gamma_2[\varphi, \Delta],$$6. The SSI penalty term for WI violation specializes to

$$\Gamma[\varphi, \Delta] = S[\varphi] + \frac{1}{2}\mathrm{Tr} \ln \Delta^{-1} + \frac{1}{2}\mathrm{Tr}\left[(\Delta_0^{-1}[\varphi]\,\Delta - 1)\right] + \Gamma_2[\varphi, \Delta],$$7

where $$\Gamma[\varphi, \Delta] = S[\varphi] + \frac{1}{2}\mathrm{Tr} \ln \Delta^{-1} + \frac{1}{2}\mathrm{Tr}\left[(\Delta_0^{-1}[\varphi]\,\Delta - 1)\right] + \Gamma_2[\varphi, \Delta],$$8 is spatial volume and $$\Gamma[\varphi, \Delta] = S[\varphi] + \frac{1}{2}\mathrm{Tr} \ln \Delta^{-1} + \frac{1}{2}\mathrm{Tr}\left[(\Delta_0^{-1}[\varphi]\,\Delta - 1)\right] + \Gamma_2[\varphi, \Delta],$$9 is inverse temperature. Thus, in the broken phase, SSI enforces the WI softly via the zero-momentum Goldstone mode.

4. Gap Equations in the Hartree–Fock Approximation

Under Hartree–Fock truncation, only "double-bubble" diagrams enter $S[\varphi]$0. The propagator ansatz in finite $S[\varphi]$1 includes an extra parameter $S[\varphi]$2: $S[\varphi]$3 For $S[\varphi]$4 (strict SI), $S[\varphi]$5, strictly enforcing the WI. The complete SSI-2PIEA includes terms for kinetic, mass, quartic, and penalty contributions, as well as Hartree–Fock diagrams. Variation yields three coupled gap equations:

• VEV equation:

$S[\varphi]$6

• Goldstone gap:

$S[\varphi]$7

• Higgs gap:

$S[\varphi]$8

All sums are renormalized using standard 2PI methods and minimal subtraction.

5. Distinct Large-Volume, Low-Temperature Limits

The SSI penalty scales as $S[\varphi]$9, so its influence depends critically on how $\Gamma_2$0 scales in the large-volume, low-temperature (infrared, IR) limit. Three limiting procedures yield fundamentally distinct behaviors:

Limit	Procedure	Phenomenology
Unimproved	$\Gamma_2$1 fixed, $\Gamma_2$2	SSI term vanishes, recovers standard 2PIEA; WI violated (e.g., $\Gamma_2$3); phase transition is second order.
Strict SI	$\Gamma_2$4 faster than $\Gamma_2$5	WI enforced strictly ($\Gamma_2$6); gap equations reproduce SI-2PIEA and Goldstone's theorem; phase transition remains second order; may lack solutions or exhibit pathology.
SSI-specific	$\Gamma_2$7, $\Gamma_2$8 fixed	Intermediate behavior: WI softly satisfied ($\Gamma_2$9), $T_{ab}^A$0; phase transition is strongly first order; broken-phase solution ceases to exist above a spinodal $T_{ab}^A$1 or for $T_{ab}^A$2; no smooth interpolation to other limits.

In all cases, the symmetric phase is unaffected by SSI because $T_{ab}^A$3 trivially there. A notable feature is the IR sensitivity of SSI: the formulation at finite volume and temperature is essential for meaningful interpolation between the different regimes. No single continuous $T_{ab}^A$4 limit encompasses all three phenomenological behaviors (Brown et al., 2016).

6. Physical and Methodological Implications

SSI provides a practical alternative to strict SI in equilibrium and potentially out-of-equilibrium scenarios. For finite $T_{ab}^A$5 and finite $T_{ab}^A$6, SSI allows the systematic study of WI violations and their phenomenological consequences. The SSI method exposes that in truncated 2PIEA treatments, exact enforcement of WIs can yield unphysical results, while their controlled violation via penalization yields richer phase structure, including strongly first-order transitions and new solution non-existence regions dependent on $T_{ab}^A$7 or $T_{ab}^A$8.

A plausible implication is that any tangible advantages of SSI, or of (S)SI methods generally, will be realized only in strictly finite-volume (or finite-temperature) applications. Attempts at perturbative expansion in $T_{ab}^A$9 or $$0 = \mathcal{W}_a^A[\varphi,\Delta] \equiv (\Delta^{-1})_{ab}T_{bc}^A\varphi_c.$$0 are hindered by singular behavior at the unimproved solution; likewise, leading-$$0 = \mathcal{W}_a^A[\varphi,\Delta] \equiv (\Delta^{-1})_{ab}T_{bc}^A\varphi_c.$$1 analysis remains trivial within this framework.

SSI generalizes to arbitrary symmetry groups and can be formulated with more elaborate kernels $$0 = \mathcal{W}_a^A[\varphi,\Delta] \equiv (\Delta^{-1})_{ab}T_{bc}^A\varphi_c.$$2, though practical implementations usually use the local, diagonal choice. Renormalization follows standard 2PI procedures with appropriate counterterms for both elementary and composite operators.

7. Relationship to Previous Work and Open Directions

Pilaftsis and Teresi introduced strict SI as a solution to WI violation, but the associated hard constraints frequently yielded no solution, especially in linear response and equilibrium truncations (Pilaftsis et al., 2013). The SSI construction by Brown and Whittingham provides a soft alternative that maintains theoretical control while improving practical tractability (Brown et al., 2016). The phenomenology revealed in the $$0 = \mathcal{W}_a^A[\varphi,\Delta] \equiv (\Delta^{-1})_{ab}T_{bc}^A\varphi_c.$$3 Hartree–Fock model establishes SSI as a framework sensitive to infrared physics, with application-dependent utility.

Open questions concern the extension of SSI to nonequilibrium dynamics, the formulation of practical algorithms for out-of-equilibrium and real-time applications, and the exploration of SSI in higher nonperturbative truncations and gauge theories. The IR sensitivity of SSI and the lack of smooth interpolation between regimes in the infinite-volume limit remain central challenges for developing robust symmetric nonperturbative quantum field theory tools.