Infra-red enhanced loops in quadratic gravity

The Big Picture: A Theory in Trouble?

Imagine physicists are trying to fix the biggest problem in modern physics: Quantum Gravity. They want a theory that explains how gravity works at the tiniest scales (like inside a black hole) without breaking math.

One popular candidate is called Quadratic Gravity. It's like a "super-charged" version of Einstein's gravity. While Einstein's theory uses simple curves (2 derivatives), this new theory adds extra complexity (4 derivatives) to make the math behave better at high energies.

The Controversy:
Recently, some researchers claimed this theory is a "golden ticket." They said that if you look at the math closely, the forces in this theory get weaker and weaker as you zoom in to higher energies. In physics jargon, they claimed the theory is "asymptotically free," meaning it becomes perfectly predictable and safe at the smallest scales.

The Authors' Verdict:
Alberto Salvio, Alessandro Strumia, and Marco Vitti (the authors of this paper) say: "Hold on a minute. That conclusion is wrong."

They argue that the "weakening force" the other researchers saw wasn't a real physical law. It was an illusion caused by how they chose to measure things and which "gauge" (mathematical coordinate system) they used. When you look at the actual, physical collisions of particles, the story changes completely.

The Core Problem: The "Map vs. Territory" Confusion

To understand their argument, let's use an analogy.

Imagine you are trying to describe the shape of a mountain.

The Territory: The actual mountain (the physical reality).
The Map: The drawing you make of it.

In physics, the "drawing" depends on your parameterization. This is just a fancy word for "how you choose to describe the variables."

You could describe the mountain's height using a ruler (Metric).
You could describe it using a balloon that inflates (Conformal).
You could describe it using a rubber sheet (Scale-factor).

The Mistake:
The previous researchers looked at their "map" (the mathematical equations) and saw that the mountain seemed to change shape depending on the energy. They thought, "Aha! The mountain itself is changing!"

The Reality:
Salvio and his team say, "No, the mountain isn't changing. You just drew the map differently."
They proved that if you change your "ruler" (parameterization) or your "compass" (gauge fixing), the apparent "running" of the forces disappears or changes completely. Since real physics shouldn't depend on which ruler you use, that "running" was fake.

The Real Discovery: The "Ghost" in the Machine

So, if the theory isn't "asymptotically free" in the way they thought, what is happening?

The authors found something even stranger. In this theory, there are "ghosts."

Analogy: Imagine a normal ball rolling down a hill (positive energy). Now imagine a "ghost ball" that rolls uphill (negative energy). In this theory, gravity has both a normal ball and a ghost ball.

When these two interact, something weird happens at high speeds:

The Illusion of Running: If you look at the math off-shell (not a real collision, just a theoretical intermediate step), it looks like the forces are changing wildly.
The Real Collision: When you calculate a real, physical collision (like two particles smashing together), the result is different. The "ghost" and the "normal" particle interfere with each other in a way that creates huge, process-dependent corrections.

The "Process-Dependent" Surprise:
In normal physics (like a car crash), if you know the speed and the car's weight, you can predict the crash. If you go faster, the physics just scales up nicely.

In Quadratic Gravity, the authors found that the result of the crash depends on exactly how the crash happens.

If two particles collide head-on, the math says one thing.
If they collide at an angle, the math says something totally different.
These differences show up as giant "logarithmic" corrections (mathematical spikes) that cannot be fixed by just saying "the force gets weaker."

It's like saying, "The speed of a car isn't just about the engine; it depends on whether the driver is wearing a red hat or a blue hat." In normal physics, the hat doesn't matter. In this theory, the "hat" (the specific details of the collision) matters a lot.

Why Does This Matter?

The Theory is Still Viable (But Weird): The paper doesn't say Quadratic Gravity is useless. It says it's not the "perfect, simple" theory the other researchers claimed. It's a complex theory where high-energy collisions are messy and depend heavily on the specific details of the event.
No "Magic Fix": You can't just tweak the math to make the theory "asymptotically free" and solve all problems. The "ghost" particles (the negative energy ones) are causing these weird effects, and they can't be ignored.
Safety Net: Interestingly, the authors note that these weird effects might actually be a good thing. Because the math gets so messy at extremely high energies (Planck scale), it might prevent particles from ever reaching those energies in the first place. It's like a "speed bump" in the universe that stops us from breaking the laws of physics.

Summary in One Sentence

The authors proved that a popular theory of gravity isn't as simple and "perfect" as previously thought; the apparent simplicity was a mathematical illusion caused by how the equations were written, and the real physics involves messy, complex interactions that depend entirely on the specific details of the collision.

1. Problem Statement

The paper addresses a controversial claim in the literature (specifically references [1–3]) regarding Quadratic Gravity (a renormalizable theory of quantum gravity in 3+1 dimensions containing $R^2$ and $R_{\mu\nu}^2$ terms).

The Claim: Previous studies suggested that logarithmically enhanced infra-red (IR) loop corrections in 4-derivative theories correspond to a physical running of couplings with external momentum ( $p$ ). Specifically, they claimed that the beta functions ( $\beta$ ) for the dimensionless couplings $f_0$ and $f_2$ depend on $\ln p$ rather than the renormalization scale $\bar{\mu}$ , rendering the theory asymptotically free in the IR.
The Conflict: Standard Quantum Field Theory (QFT) dictates that UV divergences drive the running of couplings (UV running), while IR divergences usually cancel in physical observables or are process-dependent. The authors aim to verify if the claimed "IR running" is a physical phenomenon or a mathematical artifact arising from gauge dependence and field parameterization choices.

2. Methodology

The authors employ a rigorous perturbative approach to distinguish between unphysical off-shell quantities and physical on-shell observables:

Toy Model Analysis: They first analyze a free scalar field with 2 or 4 derivatives. By performing a local field redefinition ( $\tilde{s} = s + gs^2$ ), they demonstrate that "apparent" momentum-dependent running ( $p$ -running) can be generated artificially in off-shell propagators, even in free theories. This establishes that off-shell amplitudes are not physical observables.
Gauge Fixing in Quadratic Gravity: They quantize quadratic gravity using the Faddeev-Popov procedure with a generalized gauge-fixing term containing arbitrary parameters ( $\xi_g, c_g, d_g$ ). They compute the one-loop graviton propagator in these general gauges to test the gauge dependence of the claimed IR running.
Spin-0 Sector Decomposition: To isolate the issue without the complications of spin-2 ghosts, they restrict the analysis to the spin-0 sector (governed by the $R^2$ $R^{2}$ term). They compute physical amplitudes using three different field parameterizations for the metric trace:
1. Linear trace: $g_{\mu\nu} = \eta_{\mu\nu}(1 + h/4)$
2. Conformal: $g_{\mu\nu} = e^{2\sigma}\eta_{\mu\nu}$
3. Scale factor: $g_{\mu\nu} = \Omega^2\eta_{\mu\nu}$
Physical Scattering Amplitudes: Crucially, they compute on-shell $2 \to 2$ scattering amplitudes ( $gg \to gg$ ) at one-loop order. Unlike off-shell propagators, these amplitudes must be independent of gauge and field parameterization.
2-Derivative Reformulation: They reformulate the 4-derivative theory into an equivalent theory with two 2-derivative fields (one normal scalar $\beta$ and one ghost $\alpha$ ) to clarify the origin of the logarithmic enhancements.

3. Key Contributions and Results

A. Refutation of "Physical" IR Running

The authors prove that the claimed IR running of couplings (Eq. 3 in the paper) is unphysical:

Gauge Dependence: The one-loop graviton propagator, including the terms interpreted as IR running, depends explicitly on the gauge-fixing parameters.
Parameterization Dependence: When computing the spin-0 sector using different field variables ( $h$ $h$ , $\sigma$ $σ$ , or $\Omega$ $Ω$ ), the coefficient of the $\ln p$ $ln p$ term in the off-shell propagator changes.
- In the $h$ basis, the result matches the controversial claim of [2].
- In the $\sigma$ basis, the anomalous $\ln p$ term vanishes.
- In the $\Omega$ basis, a different coefficient appears.
Conclusion: Since physical observables cannot depend on arbitrary choices of field coordinates or gauge, the "IR running" is an artifact of using off-shell quantities.

B. Physical On-Shell Amplitudes

The authors compute physical $2 \to 2$ scattering amplitudes. Their results show:

Process Dependence: The amplitudes exhibit genuine, log-enhanced IR corrections. However, these corrections are process-dependent (they differ for $00 \to 00$ vs. $00 \to M_0 M_0$ scattering).
No Universal Running: Because the corrections depend on the specific scattering process, they cannot be absorbed into a universal running of the coupling constants ( $f_0$ or $f_2$ ).
Breakdown of Standard EFT Intuition: In standard QFT, high-energy scattering is approximated by tree-level diagrams with couplings renormalized at the energy scale. In quadratic gravity, this approximation fails; the loop corrections contain large, process-dependent logarithms that are not captured by simple coupling renormalization.

C. The Origin: Singular $M_0 \to 0$ Limit

The paper identifies the root cause of these unusual IR enhancements:

Ghost States: Quadratic gravity contains a massive spin-0 ghost (and spin-2 ghost).
Singular Diagonalization: When reformulating the theory into two 2-derivative fields (a normal scalar and a ghost), the transformation required to diagonalize the mass and kinetic terms involves an $SO(1,1)$ boost.
The Singularity: As the mass of the ghost ( $M_0$ ) approaches zero, this boost parameter becomes singular (infinite). Consequently, the limit $M_0 \to 0$ is singular. Small mass terms, which kinematically distinguish the two states, play a critical role even at energies $E \gg M_0$ .
Mechanism: The "soft" mass term breaks the $SO(1,1)$ symmetry in a "hard" way, leading to the emergence of the unusual logarithmic enhancements in the scattering amplitudes.

D. Effective Action and Positivity

Tree-Level EFT: The authors show that the tree-level effective action of pure quadratic gravity (integrating out massive modes) is exactly Einstein Gravity.
Positivity Bounds: Despite the presence of ghosts in the full theory, the tree-level EFT satisfies standard positivity bounds on non-renormalizable operators. The ghosts only become visible at the loop level or when matter is coupled.

4. Significance and Implications

Asymptotic Freedom Claim Debunked: The paper refutes the claim that quadratic gravity is asymptotically free due to IR effects. The beta functions derived from off-shell propagators in specific gauges are not physical. The theory's UV behavior remains governed by the standard gauge-independent UV beta functions (Eq. 2), where $f_0$ is not asymptotically free.
New QFT Phenomenon: The work highlights a novel feature of 4-derivative theories (and theories with ghosts): Process-dependent IR enhancements. These are genuine physical effects that cannot be resummed into a simple running coupling, challenging the standard intuition of how high-energy scattering behaves in such theories.
Perturbative Validity: The authors argue that these large logarithms do not invalidate perturbation theory in quadratic gravity. Tree-level IR enhancements become large only near the Planck scale. At sub-Planckian energies, the theory remains calculable, and the "screening" of trans-Planckian physics by soft graviton emission prevents the loop corrections from becoming uncontrollable.
Methodological Warning: The paper serves as a critical warning against interpreting off-shell propagator corrections as physical running couplings, especially in gauge theories and theories with higher derivatives where field reparameterization invariance is subtle.

In summary, Salvio, Strumia, and Vitti demonstrate that while quadratic gravity exhibits unique and strong IR logarithmic corrections in physical scattering amplitudes, these are process-dependent artifacts of the ghost sector's singular mass limit, not evidence of a new physical running of couplings that would render the theory asymptotically free.