⚛️ general relativity

Removing spurious degrees of freedom from EFT of gravity

This paper demonstrates that by applying an action-based procedure to remove spurious higher-derivative degrees of freedom from general relativity supplemented with cubic Riemann terms, the resulting effective theory of gravity can be reformulated as a minimally modified gravity model that propagates only the massless spin-2 graviton while exhibiting a preferred frame at short distances.

Original authors: Dražen Glavan, Shinji Mukohyama, Tom Zlosnik

Published 2026-07-31

📖 3 min read🧠 Deep dive

CC BY 4.0

Original authors: Dražen Glavan, Shinji Mukohyama, Tom Zlosnik

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the universe as a giant, invisible trampoline made of spacetime. When you place a heavy bowling ball (like a star) on it, the fabric curves, and that curvature is what we feel as gravity. For nearly a century, our best description of this trampoline has been Albert Einstein's General Relativity. It's a brilliant theory that works perfectly for most things we see, from falling apples to orbiting planets. But scientists suspect that if we zoom in really, really close—down to the tiniest, most energetic scales imaginable—Einstein's rules might start to wobble. To fix this, physicists use a tool called "Effective Field Theory." Think of this as a "patch kit" for the universe. It doesn't throw away Einstein's theory; instead, it adds tiny, extra instructions to the rulebook to account for the weirdness of the super-small world.

However, there's a catch. When you add these extra instructions to the math, you often accidentally introduce "ghosts." These aren't scary spirits, but mathematical glitches that act like extra, invisible particles popping into existence out of nowhere. In the real world, we only see one kind of gravity particle (the graviton), but these mathematical ghosts would suggest there are more. If we take these equations at face value, they predict a universe that is unstable and full of nonsense. The big question for physicists is: How do we keep the helpful "patches" that fix the high-energy problems without inviting these unwanted ghosts into the party?

This is exactly the puzzle tackled by Dražen Glavan, Shinji Mukohyama, and Tom Zlosnik in their paper, "Removing spurious degrees of freedom from EFT of gravity." They focus on a specific, tricky patch involving a term called "Riemann-cubed." In the language of gravity math, the Riemann tensor describes how spacetime curves. Squaring it or cubing it creates complex corrections for extreme conditions. The authors show that if you just write down these cubic terms, your equations scream that there are extra, ghostly degrees of freedom (extra ways the universe can wiggle). But, they argue, this is a trick of the math, not a feature of reality.

Using a clever mathematical procedure called "action-based derivative reduction," the team demonstrates how to surgically remove these ghosts. They treat the extra terms not as new, independent rules, but as subtle adjustments to the existing ones. By doing this, they strip away the extra wiggles and are left with a clean theory that still has only the two physical polarizations of the graviton (the two ways a gravitational wave can shake). The result is a theory that looks very much like Einstein's General Relativity but with a twist: at very short distances, it behaves as if time has a preferred direction, breaking the perfect symmetry between space and time in a specific way. This new theory belongs to a class called "minimally modified gravity."

The authors suggest that this approach offers a promising way to understand how the universe might behave under the influence of unknown, ultra-high-energy physics without breaking the fundamental rules of stability. They prove that their method works for this specific cubic correction, showing that the "ghosts" were never real to begin with—they were just artifacts of how the math was written. While they don't claim to have solved the entire mystery of quantum gravity, they provide a robust blueprint for how to clean up these effective theories, ensuring that the predictions we make about the cosmos remain grounded in a stable, sensible reality.

Technical Summary: Removing Spurious Degrees of Freedom from EFT of Gravity

Problem Statement
In the Effective Field Theory (EFT) approach to gravity, the Einstein-Hilbert action is supplemented by higher-order curvature terms (e.g., Riemann tensor cubed) to encode ultraviolet (UV) corrections. While the underlying UV-complete theory presumably propagates only the massless spin-2 graviton, a naive truncation of the local derivative expansion results in higher-derivative equations of motion. These equations typically introduce "spurious" degrees of freedom (DOF) beyond the two physical polarizations of the graviton. These extra modes are often associated with Ostrogradsky instabilities, rendering the truncated theory unstable and conceptually inconsistent with the goal of EFT, which is to describe low-energy physics using only the existing low-energy DOFs.

Standard perturbative approaches, which treat corrections as a power series around a background solution, avoid introducing new DOFs but are restrictive; they preclude exact solutions or resummations that might be necessary in certain regimes (e.g., long-time evolution). Conversely, treating the higher-derivative terms as fundamental leads to unphysical instabilities. The paper addresses the challenge of explicitly removing these spurious DOFs from the local effective action to derive a healthy, second-order-in-time theory that captures UV effects without introducing new propagating modes.

Methodology
The authors apply an action-based derivative reduction procedure grounded in the canonical (Hamiltonian) formalism, extending the algorithm proposed in [17] (building on [18, 19]) to theories with local symmetries (diffeomorphism invariance). The specific model considered is General Relativity supplemented by a cubic Riemann tensor term:
$S[g_{\mu\nu}] = \int d^4x \sqrt{-g} \left[ \frac{R - 2\Lambda}{\kappa^2} + \alpha \kappa^2 R_{\mu\nu}^{\rho\sigma} R_{\rho\sigma}^{\alpha\beta} R_{\alpha\beta}^{\mu\nu} \right]$
where $\alpha \ll 1$ is a small coupling constant.

The procedure involves the following steps:

ADM Decomposition: The metric is decomposed into the lapse $N$ , shift $N^i$ , and spatial metric $g_{ij}$ . The Riemann-cubed term is decomposed into ADM variables, revealing dependence on second time derivatives of the metric (contained in the tensor $F_{ij}$ ).
Extended Action: An extended action is constructed by promoting time derivatives to independent velocity fields ( $K_{ij}, F_{ij}$ ) and introducing Lagrange multipliers ( $\pi_{ij}, \rho_{ij}$ ) to enforce definitions.
Perturbative Analyticity Assumption: A crucial assumption is introduced: all fields must be formally analytic functions of the coupling constant $\alpha$ on-shell. This prevents the division by $\alpha$ that would otherwise be required to solve for auxiliary variables directly, effectively treating the higher-derivative term as a perturbative correction rather than a fundamental modification.
Constraint Analysis: The requirement of analyticity generates effective second-class constraints. By solving the equations of motion order-by-order in $\alpha$ , the authors eliminate the dynamical variables associated with spurious DOFs (specifically solving for $\rho_{ij}$ and $F_{ij}$ in terms of canonical variables $g_{ij}$ and $\pi_{ij}$ ).
Canonical Simplification: The resulting reduced action is simplified via field redefinitions (shifts in $\pi_{ij}$ and $g_{ij}$ ) to restore canonical Poisson brackets, followed by redefinitions of the lapse and shift to simplify the Hamiltonian density.

Key Contributions and Results

Reduction to Second Order: The procedure successfully reduces the order of time derivatives in the equations of motion from higher-order to second-order, eliminating the spurious degrees of freedom associated with the cubic curvature term.
Minimally Modified Gravity: The resulting reduced theory is identified as a member of the "minimally modified gravity" class. It preserves:
1. Spatial diffeomorphism invariance.
2. The number of propagating degrees of freedom (exactly two, the massless spin-2 graviton).
3. The on-shell first-class constraint structure of General Relativity (GR).
Deformation of Time Diffeomorphisms: While spatial symmetries are preserved, the reduction procedure selects a preferred frame at short distances, deforming time diffeomorphisms. This distinguishes the theory from standard GR.
Non-Equivalence to GR: The authors demonstrate that the correction to the Hamiltonian constraint cannot be absorbed into further local field redefinitions (canonical transformations). Specifically, terms involving six factors of canonical momenta in the reduced Hamiltonian cannot be generated by the transformation of a local scalar functional. Thus, the theory is physically distinct from GR and belongs to the Type-II class of minimally modified gravity.
Constraint Algebra: The algebra of the reduced constraints is computed to first order in $\alpha$ . The constraints remain first-class, and their Poisson brackets vanish on-shell, ensuring the consistency of the theory and the preservation of the correct number of DOFs.
Gravitational Waves: In Appendix B, the propagation of gravitational waves on a maximally symmetric background is analyzed. To order $\alpha$ , the wave speed and amplitude can be redefined via field transformations to match GR, suggesting that linear perturbations in vacuum are degenerate with GR up to field redefinitions, though matter coupling may break this degeneracy.

Significance and Claims
The paper claims to provide a systematic method for extracting reliable predictions from higher-derivative local effective theories of gravity by explicitly excising spurious degrees of freedom. The authors argue that the resulting theory offers a consistent effective description of gravity that captures UV corrections (via the cubic term) while maintaining the physical integrity of the low-energy sector (only the graviton propagates).

The work suggests that "minimally modified gravity" theories, which were previously introduced via classical considerations, can be understood as the natural effective field theory limit of gravity when higher-derivative corrections are treated perturbatively and spurious modes are removed. The authors note that while the reduction is performed perturbatively (to order $\alpha$ ), the procedure is conceptually extendable to higher orders, though this remains an open question. They also highlight that the presence of matter fields could alter the coupling structure, a topic requiring further investigation. The paper does not claim to solve the UV completion of gravity but rather to provide a consistent low-energy effective framework that avoids the pathologies of naive truncation.

More like this