Compare · A Theory of Everything

Higher-Derivative Gravity Extensions vs Asymptotic Safety (Quantum Einstein Gravity)

← Back to Higher-Derivative Gravity Extensions

Asymptotic Safety· within family

	Higher-Derivative Gravity Extensions Frontier	Asymptotic Safety (Quantum Einstein Gravity) Frontier
Proposed	1977	1979 / 1998
Key figures	Kellogg Stelle, Alessandro Codello, Roberto Percacci, Benjamin Knorr, Frank Saueressig	Steven Weinberg, Martin Reuter
In one sentence	Stelle showed in 1977 that gravitational theories with curvature-squared terms (R-squared, Ricci-squared, Weyl-squared) added to the Einstein-Hilbert action are perturbatively renormalizable but contain ghosts: negative-norm states that ruin probability conservation. Codello-Percacci 2006 showed that within asymptotic safety's non-perturbative framework, fixed points exist for the higher-derivative couplings too, potentially resolving the ghost problem. Modern work on form factors by Knorr, Ripken, and Saueressig is the current state of the art.	Weinberg proposed in 1979 that gravity's ultraviolet behavior is tamed not by extra physics but by a non-trivial fixed point: the gravitational couplings approach a finite value at high energies rather than blowing up. Reuter made the proposal calculable in 1998 by writing down a functional renormalization-group equation that tracks how the couplings flow with energy. The past three decades have been about increasingly realistic checks that the fixed point really exists.
Predictions	Fixed points in the renormalization-group flow exist for higher-derivative gravitational operators (R-squared, Ricci-squared, Weyl-squared couplings), not just for Newton's constant and the [[cosmological constant]] Curvature-dependent form factors, the functions parameterizing the full quantum corrections to the gravitational action, approach scaling forms at the fixed point that can be computed within truncation The classical ghost states of perturbative higher-derivative gravity are absent (or rendered harmless) in the non-perturbative asymptotic-safety completion Specific dimensionless ratios involving the higher-derivative couplings at the fixed point should match between different truncation schemes; agreement is consistency evidence, not a proof of the underlying theory	Gravitational couplings approach a non-trivial fixed point at energies near the Planck scale; this is the central testable structural prediction of the framework The theory remains predictive at arbitrarily high energies without needing extra fields or strings; quantum gravity is a self-contained sector Specific dimensionless ratios at the fixed point can be computed (e.g. the product of the cosmological constant and Newton's constant in fixed-point units); these ratios should be consistent across different truncations and matter contents Newtonian gravity is recovered at low energies as the infrared limit of the renormalization-group flow from the fixed point; this is a consistency requirement, not a free prediction
Where it breaks	The truncation convergence problem hits this variant especially hard: higher-derivative truncations are larger than the Einstein-Hilbert case, but the operator space is also bigger, so it is not obvious the convergence picture improves rather than just becoming more complicated The claim that ghosts are resolved by the non-perturbative completion is a hopeful interpretation of fixed-point evidence rather than a proof; a constructive demonstration that the non-perturbative theory is unitary is still missing Higher-derivative gravity theories are notoriously difficult to formulate causally; standard Cauchy-problem analyses produce instabilities that the perturbative framework cannot resolve The form-factor program produces consistent fixed-point structures, but the physical interpretation of those structures (what kind of theory they actually describe at high energies) is less developed than the technical calculations themselves	Truncation convergence is unproven (see family-level sharedObjections for full statement). The strongest objection to the program is structural: no proof exists that the truncations converge to the true theory Most calculations are Euclidean; the Lorentzian carry-over is contested (see Lorentzian variant for the active research line attempting to address this) Asymptotic safety has no distinctive low-energy prediction confirmed experimentally beyond the conditional Higgs result, leaving the framework's status similar to other quantum-gravity proposals Some authors (Donoghue 2020 and others) have argued that asymptotic safety as formulated may not survive once non-perturbative effects beyond the renormalization-group truncation are properly included; this is a sharpened form of the truncation-convergence concern
Key unresolved problem	The bad-probabilities problem: older versions of this kind of gravity produce ghost states that imply negative probabilities, and no one has yet proven the asymptotic-safety version is free of them, that it stays unitary.	The approximation problem: the key result comes from cutting the equations short to make them solvable, and no one has proven that the full untruncated calculation would settle on the same fixed point rather than wandering off.
Reader vote	No votes yet	No votes yet

Higher-Derivative Gravity Extensions

1977 · Frontier

Asymptotic Safety (Quantum Einstein Gravity)

1979 / 1998 · Frontier

Proposed

1977

1979 / 1998

Key figures

Kellogg Stelle, Alessandro Codello, Roberto Percacci, Benjamin Knorr, Frank Saueressig

Steven Weinberg, Martin Reuter

In one sentence

Stelle showed in 1977 that gravitational theories with curvature-squared terms (R-squared, Ricci-squared, Weyl-squared) added to the Einstein-Hilbert action are perturbatively renormalizable but contain ghosts: negative-norm states that ruin probability conservation. Codello-Percacci 2006 showed that within asymptotic safety's non-perturbative framework, fixed points exist for the higher-derivative couplings too, potentially resolving the ghost problem. Modern work on form factors by Knorr, Ripken, and Saueressig is the current state of the art.

Weinberg proposed in 1979 that gravity's ultraviolet behavior is tamed not by extra physics but by a non-trivial fixed point: the gravitational couplings approach a finite value at high energies rather than blowing up. Reuter made the proposal calculable in 1998 by writing down a functional renormalization-group equation that tracks how the couplings flow with energy. The past three decades have been about increasingly realistic checks that the fixed point really exists.

Predictions

Fixed points in the renormalization-group flow exist for higher-derivative gravitational operators (R-squared, Ricci-squared, Weyl-squared couplings), not just for Newton's constant and the [[cosmological constant]]
Curvature-dependent form factors, the functions parameterizing the full quantum corrections to the gravitational action, approach scaling forms at the fixed point that can be computed within truncation
The classical ghost states of perturbative higher-derivative gravity are absent (or rendered harmless) in the non-perturbative asymptotic-safety completion
Specific dimensionless ratios involving the higher-derivative couplings at the fixed point should match between different truncation schemes; agreement is consistency evidence, not a proof of the underlying theory

Gravitational couplings approach a non-trivial fixed point at energies near the Planck scale; this is the central testable structural prediction of the framework
The theory remains predictive at arbitrarily high energies without needing extra fields or strings; quantum gravity is a self-contained sector
Specific dimensionless ratios at the fixed point can be computed (e.g. the product of the cosmological constant and Newton's constant in fixed-point units); these ratios should be consistent across different truncations and matter contents
Newtonian gravity is recovered at low energies as the infrared limit of the renormalization-group flow from the fixed point; this is a consistency requirement, not a free prediction

Where it breaks

The truncation convergence problem hits this variant especially hard: higher-derivative truncations are larger than the Einstein-Hilbert case, but the operator space is also bigger, so it is not obvious the convergence picture improves rather than just becoming more complicated
The claim that ghosts are resolved by the non-perturbative completion is a hopeful interpretation of fixed-point evidence rather than a proof; a constructive demonstration that the non-perturbative theory is unitary is still missing
Higher-derivative gravity theories are notoriously difficult to formulate causally; standard Cauchy-problem analyses produce instabilities that the perturbative framework cannot resolve
The form-factor program produces consistent fixed-point structures, but the physical interpretation of those structures (what kind of theory they actually describe at high energies) is less developed than the technical calculations themselves

Truncation convergence is unproven (see family-level sharedObjections for full statement). The strongest objection to the program is structural: no proof exists that the truncations converge to the true theory
Most calculations are Euclidean; the Lorentzian carry-over is contested (see Lorentzian variant for the active research line attempting to address this)
Asymptotic safety has no distinctive low-energy prediction confirmed experimentally beyond the conditional Higgs result, leaving the framework's status similar to other quantum-gravity proposals
Some authors (Donoghue 2020 and others) have argued that asymptotic safety as formulated may not survive once non-perturbative effects beyond the renormalization-group truncation are properly included; this is a sharpened form of the truncation-convergence concern

Key unresolved problem

The bad-probabilities problem: older versions of this kind of gravity produce ghost states that imply negative probabilities, and no one has yet proven the asymptotic-safety version is free of them, that it stays unitary.

The approximation problem: the key result comes from cutting the equations short to make them solvable, and no one has proven that the full untruncated calculation would settle on the same fixed point rather than wandering off.

Reader vote

No votes yet