---
title: Fourth-Order Gravity Theories
url: https://www.emergentmind.com/topics/fourth-order-gravity-theories
type: topic
---

# Fourth-Order Gravity Theories

Fourth-order gravity theories constitute a broad class of metric theories that augment or replace the Einstein–Hilbert action with curvature-squared and related higher-derivative invariants, resulting in field equations of fourth order in the metric. These models have been considered for their potential role in ultraviolet (UV) completions of general relativity (GR), for implications in classical and quantum cosmology, and as possible explanations for astrophysical phenomena without invoking dark matter or dark energy.

## 1. Action and Field Equations in Fourth-Order Gravity

The foundational action for fourth-order gravity typically generalizes the Einstein–Hilbert action to include quadratic curvature invariants. The canonical form is
\[
S = \int d^4x\,\sqrt{-g}\Big[ R + a\,R^2 + b\,R_{\mu\nu}R^{\mu\nu} \Big] + S_{\rm matter}
\]
where \(R\) is the Ricci scalar, \(R_{\mu\nu}\) the Ricci tensor, and \(a\), \(b\) are coupling constants with dimensions of length squared [1108.1665]. The most general form may also include Riemann tensor squared and higher-order terms,
\[
S = \int d^4x\,\sqrt{-g}\,f\big(R, R_{\mu\nu}R^{\mu\nu}, R_{\mu\nu\rho\sigma}R^{\mu\nu\rho\sigma}\big) + S_{\rm m}
\]
with associated field equations containing up to fourth derivatives of the metric [1007.1917, 1501.02187, 2311.07246].

Variation with respect to \(g_{\mu\nu}\) yields field equations that, in compact notation for the quadratic theory, cast as:
\[
G_{\mu\nu} = 8\pi G\,\left( T^{\rm matter}_{\mu\nu} + T^{\rm vac}_{\mu\nu} \right)
\]
where \(T^{\rm vac}_{\mu\nu}\) subsumes the higher-derivative contributions, with principal terms:
\[
T^{\rm vac}_{\mu\nu} = -(2a+b) [g_{\mu\nu}\Box R - \nabla_\mu\nabla_\nu R] - 2b [ \Box G_{\mu\nu} + 2R_{\mu\alpha\nu\beta}G^{\alpha\beta} ] + \dots
\]
[1108.1665]. These field equations exhibit scalar and massive spin-2 degrees of freedom beyond the massless spin-2 graviton of GR.

## 2. Newtonian and Weak-Field Limit: Emergence of Yukawa Corrections

In the static, spherically symmetric, weak-field regime (Newtonian approximation), the field equations linearize and decouple into ordinary differential equations for scalar and spin-2 modes. For the quadratic theory,
\[
g_{\mu\nu} = \eta_{\mu\nu} + h_{\mu\nu},\quad |h_{\mu\nu}|\ll 1
\]
the trace and spin-2 sector reduce to (neglecting pressure relative to density):
- Scalar (trace) mode:
\[
(6a+2b)\nabla^2 T - T = -\rho_{\mathrm{mat}}
\]
- Spin-2 combination:
\[
b \nabla^2 (\rho+3p) + (\rho+3p) = 2\rho_{\mathrm{mat}}
\]
where \(\rho = T^0_0\), \(T = T^\mu_\mu\).

The solutions outside a spherical source (\(r > R\), \(R\) = body radius) are sums of Yukawa-type terms with characteristic lengths:
\[
\lambda_1 = \sqrt{6a+2b}, \quad \lambda_2 = \sqrt{-b}
\]
\[
T(r) \sim \frac{e^{-r/\lambda_1}}{r},\qquad (\rho+3p)(r) \sim \frac{e^{-r/\lambda_2}}{r}
\]
Thus, the gravitational acceleration and potential for \(r > R\) become [1108.1665, 1007.1917, 2311.07246]:
\[
g(r) = -\frac{GM}{r^2} \left\{ 1 + \frac{1}{3}(1 + r/\lambda_1)e^{-(r-R)/\lambda_1} - \frac{4}{3}(1 + r/\lambda_2)e^{-(r-R)/\lambda_2} \right\}
\]
\[
\Phi(r) = -\frac{GM}{r} \left[ 1 + \alpha\,e^{-r/\lambda_1} + \beta\,e^{-r/\lambda_2} \right],\quad \alpha = +\frac{1}{3},\;\beta = -\frac{4}{3}
\]
showing one attractive and one repulsive Yukawa correction to the Newtonian term.

## 3. Physical and Experimental Constraints on Characteristic Ranges

Precision laboratory tests—torsion balances and Cavendish-type experiments—impose stringent bounds on any non-Newtonian, Yukawa gravitational corrections with order-unity couplings. Specifically [1108.1665]:
\[
\lambda_1 = \sqrt{6a+2b} \lesssim \text{a few mm},\quad \lambda_2 = \sqrt{-b} \lesssim \text{a few mm}
\]
implying
\[
a,\,|b| \lesssim 10^{-6}\,\rm m^2
\]
as larger ranges would result in detectable deviations from Newton's law at sub-millimeter scales, which are not observed.

## 4. Decoupling and Astrophysical/Cosmological Implications

Given the above experimental bounds, the two Yukawa modes are highly short-range:
- For laboratory-scale bodies (\(R\) up to meters): For \(a \sim -b \ll R^2\), nonlinear corrections are negligible and Newtonian gravity is recovered at all relevant distances.
- For larger \(a,b\): If \(a\sim -b \gg R^2\), the gravitational field is exponentially suppressed near the body and only asymptotically becomes Newtonian at \(r \gtrsim \sqrt{a}\sim\sqrt{-b}\). However, this regime is experimentally excluded for macroscopic \(a,\,b\).
- For astrophysical and cosmological scales (\(R_{\rm star}, R_{\rm gal}, R_H \gg \text{mm}\)): Correction terms have died out far below the relevant radii, and the gravitational field is indistinguishable from pure Newtonian or Einstein gravity [1108.1665].

Thus, fourth-order corrections of the \(aR^2 + b R_{\mu\nu}R^{\mu\nu}\) type have negligible impact on the structure and dynamics of stars, galaxies, or the universe’s large-scale evolution since their effect is confined to sub-millimeter scales by laboratory constraints.

## 5. Mathematical Structure and Decoupling of Modes

The mass parameters for the additional degrees of freedom are determined algebraically by the Lagrangian coefficients:
\[
m_0^2 = \frac{1}{6a+2b},\quad m_2^2 = -\frac{1}{b}
\]
with positivity requirements for real, exponential decay [1007.1917, 1501.02187]. In the general action including all possible quadratic invariants,
\[
S = \int d^4x\,\sqrt{-g}\left(a_1 R + a_2 R^2 + a_3 R_{\mu\nu}R^{\mu\nu} + a_4 R_{\mu\nu\rho\sigma}R^{\mu\nu\rho\sigma}\right)
\]
only two Yukawa scales survive at the Newtonian level, due to the Gauss–Bonnet theorem rendering the Riemann-squared term dynamically redundant in four dimensions [1007.1917].

## 6. Theoretical Extensions and Limitations

While some modified gravity models with Yukawa corrections have been invoked to explain galactic rotation curves or alternative dark matter/energy phenomenology, the required scale for the onset of new physics (kiloparsec–gigaparsec) is ruled out in these simple \(R^2 + R_{\mu\nu}R^{\mu\nu}\) models by sub-millimeter laboratory data [1108.1665]. Scalar–tensor extensions, additional dynamical fields, or non-local constructions are required for long-range modifications compatible with experiment.

A plausible implication is that, for any general theory of the \(R+a R^2 + b R_{\mu\nu} R^{\mu\nu}\) type, all observable effects on macroscopic gravitational phenomena are excluded by current small-scale experiments, so that physically relevant modifications to gravity at larger scales must arise via different mechanisms.

---

**References**  
- "Constraints on fourth order generalized f(R) gravity" [1108.1665]  
- "The most general fourth order theory of Gravity at low energy" [1007.1917]  
- "Post-Minkowskian Limit and Gravitational Waves solutions of Fourth Order Gravity: a complete study" [1501.02187]  
- "Multipole expansion of the gravitational field in a general class of fourth-order theories of gravity and the application in gyroscopic precession" [2311.07246]

Source: https://www.emergentmind.com/topics/fourth-order-gravity-theories