---
title: 'EFTofDE: Effective Field Theory of Dark Energy'
url: https://www.emergentmind.com/topics/effective-field-theory-of-dark-energy-eftofde
type: topic
---

# EFTofDE: Effective Field Theory of Dark Energy

The effective field theory of dark energy (EFTofDE) provides a systematic, model-independent parametrization for the low-energy phenomenology of cosmological acceleration, encompassing both canonical quintessence, general Horndeski, and beyond-Horndeski single-scalar extensions of general relativity. The EFTofDE trades explicit model Lagrangians for a finite set of time-dependent operator coefficients in the action, encoding all allowed departures from ΛCDM (at the level of background evolution and perturbations) consistent with symmetries and stability.

## 1. Unitary-Gauge Action and Background Evolution

In the EFTofDE, the extra scalar field (e.g., a dark energy or modified gravity scalar) is “gauge-fixed” such that perturbations coincide with the metric time slicing (unitary gauge). The most general local action up to quadratic order in metric perturbations and two derivatives can be written as:
\[
S = \int d^4x \sqrt{-g} \Bigg[ \frac{M_*^2(t)}{2} R - \Lambda(t) - c(t) g^{00}
+ \frac{M_2^4(t)}{2} (\delta g^{00})^2
- \frac{\bar{m}_1^3(t)}{2} \delta g^{00}\delta K
- \frac{\bar{M}_2^2(t)}{2} (\delta K)^2
- \frac{\bar{M}_3^2(t)}{2} (\delta K^{\mu}_{\;\;\nu}\delta K^{\nu}_{\;\;\mu})
+ \frac{\mu_1^2(t)}{2} \delta g^{00} \delta R^{(3)} + ...
\Bigg] + S_m
\]
with definitions:
- $M_*^2(t)$: effective (possibly time-varying) Planck mass,
- $\Lambda(t)$: time-dependent vacuum (cosmological constant) energy,
- $c(t)$: controls the kinetic mixing between gravity and the scalar,
- $M_2^4$, $\bar{m}_1^3$, $\bar{M}_{2,3}^2$, $\mu_1^2$, ...: perturbative operators,
- $S_m$: minimally coupled matter sector ($\rho_m$, $\rho_r$, ...).

The background evolution, for spatially flat FRW, is governed by modified Friedmann equations,
\[
\begin{aligned}
3M_*^2 H^2 &= \rho_m + \Lambda + c - 3H \dot{M}_*^2 \\
(2\dot{H} + 3H^2) M_*^2 &= -p_m + \Lambda - c - 2H\dot{M}_*^2 - \ddot{M}_*^2
\end{aligned}
\]
where the matter stress energy evolves separately.

The entire homogeneous sector is thus specified by $\{M_*^2(t),\, \Lambda(t),\, c(t)\}$, reducing a large landscape of DE/MG models to only three free functions controlling $H(a)$ for a fixed matter content [1310.6026][1211.7054][1907.03150].

## 2. Operator Basis and the $\alpha$-Parameterization

For linear perturbations, the Bellini–Sawicki $\alpha$-basis renders the observable content transparent:
\[
\begin{aligned}
\alpha_M &\equiv \frac{1}{H M_*^2} \frac{dM_*^2}{dt} \quad\text{(evolution of effective Planck mass)} \\
\alpha_K &\equiv \frac{2c+4M_2^4}{M_*^2 H^2} \quad\text{(scalar kinetic term)} \\
\alpha_B &\equiv -\frac{\bar{m}_1^3}{2M_*^2 H} \quad\text{(kinetic mixing/"braiding")} \\
\alpha_T &\equiv \frac{\bar{M}_3^2-\bar{M}_2^2}{M_*^2} \quad\text{(excess tensor speed)} \\
\alpha_H &\equiv ... \qquad\text{(beyond-Horndeski)}
\end{aligned}
\]
In pure Horndeski, only four $\alpha$-functions are independent; general beyond-Horndeski extensions have up to seven [1512.06180][1907.03150]. The mapping between the original operator coefficients and $\{\alpha_i\}$ is invertible.

For canonical scalar-tensor models, e.g., quintessence,
- $\alpha_K > 0$,
- $\alpha_B=\alpha_M=\alpha_T=\alpha_H=0$.

Screening classes (e.g., DGP, Galileon) activate more operators [1312.6111][1706.02719].

## 3. Stability, Causality, and Continuity Structure

Ghost and gradient stability require:
- positive-definite kinetic matrix for the scalar sector,
- $c_s^2 > 0$ for the scalar sound speed (no gradient instability), typically,
\[
Q_s = M_*^2 \frac{\alpha_K + 6\alpha_B^2}{(1+\alpha_B)^2} > 0, \quad
c_s^2 = \mathrm{complicated\ function\ of}\ \{\alpha_i\} > 0,
\]
with explicit forms in [1907.03150][1512.06180][1810.05225].

The choice of parameterization can impact both stability and the physical interpretability of the background sector. The continuity-equation-compatible (CEC) basis [2507.18526] eliminates spurious non-conservation by directly matching the dark energy density and pressure to combinations of the original $\{c,\Lambda\}$, enforcing
\[
\dot\rho_{\rm de} + 3H(\rho_{\rm de}+P_{\rm de}) = 0.
\]
This basis is advantageous for numerical implementation and connecting to observational quantities (e.g., input into EFTCAMB).

## 4. Linear Perturbations and Phenomenological Functions

On linear, subhorizon scales, two metric potentials $\Psi$ and $\Phi$ obey modified Poisson and anisotropy equations:
\[
-\frac{k^2}{a^2} \Phi = 4\pi G\ \mu(a,k)\ \rho_m \delta_m, \qquad \eta(a,k) = \frac{\Phi}{\Psi},
\]
where $\mu(a,k)$ (effective Newton constant), $\Sigma(a,k)$ (lensing response), and $\eta(a,k)$ (metric slip) are algebraic functions of the $\alpha$-coefficients [1903.08088][1907.03150][2512.03314].

The growth of structure probes $\mu(a)$ via redshift-space distortions and $f\sigma_8$, while weak lensing constrains $\Sigma(a,k)$.

Post-GW170817, $\alpha_T \simeq 0$ is enforced for luminal gravitational wave propagation, ruling out large sections of the quartic/quintic Horndeski parameter space unless dynamically suppressed [1903.08088]. In this regime, $c_T^2=1$ and only $\{\alpha_K, \alpha_B, \alpha_M\}$ remain active for scalar phenomenology.

## 5. Nonlinear Extensions and Screening

To extend EFTofDE to mildly non-linear scales, all higher-order operators in the action must be included, e.g., $(\delta g^{00})^3$, $(\delta K)^3$, and terms mixing the lapse, curvature, and extrinsic curvature (see full enumeration in [1706.02719]). Screening mechanisms such as Vainshtein arise from such non-linear operators and are critical for viable modified gravity models. The action up to quartic order can be written in ADM language and mapped to covariant Horndeski or beyond-Horndeski Lagrangians.

N-body implementations of these nonlinear extensions—e.g., ECOSMOG-EFT and PySCo-EFT—solve the full coupled system for the metric potentials and scalar, including nonlinear screening terms, and allow for robust predictions for the matter power spectrum boost, matching analytic theory to sub-percent accuracy for $k \lesssim 10h/$Mpc in the parameter regime allowed by current data [2604.15434].

## 6. Quantum-Kinetic Dark Energy (QKDE): Minimal Conservative Realization

Quantum-Kinetic Dark Energy [2603.14716] realizes the most conservative sector of EFTofDE: the only deviation from ΛCDM is a time-dependent scalar kinetic normalization $K(\chi)$, with
- $\alpha_K > 0$ (ensuring scalar stability),
- $\alpha_B = \alpha_M = \alpha_T = \alpha_H = 0$ (no braiding, Planck-mass running, or tensor speed shift).

In this case,
- $c_s^2=1$, tensors always propagate at the speed of light,
- All linear response functions $\mu(a,k) = \Sigma(a,k) = 1$, $\eta(a,k)=0$: scalar and tensor cosmological phenomenology is unmodified,
- The only observable effect is via the background $H(a)$ and growth $D(a)$: all growth/lensing is as in general relativity for a given expansion history.

Any deviation from these signatures in large-scale structure/growth or gravitational wave propagation would immediately falsify the QKDE scenario.

The action is
\[
S = \int d^4x \sqrt{-g} \left[\frac12 M_\mathrm{pl}^2 R + K(\chi) X - V(\Phi) \right] + S_m + S_\chi,
\]
where $K(\chi) > 0$, $\chi$ is a covariant "clock" field, and $X = -\frac12 g^{\mu\nu} \partial_\mu \Phi\, \partial_\nu \Phi$. The background evolution and perturbative sector reduce exactly to GR plus a minimally coupled scalar with a time-dependent kinetic normalization.

Explicit forms for $K(\chi)$, including curvature-coupled or phenomenological running, provide UV- or phenomenologically-motivated models that can be tested against $H(z)$ and $f\sigma_8$ data.

## 7. Model-Independent Reconstruction and Observational Tests

Reconstruction of the three background EFT functions from cosmic chronometer, SNe, BAO, or CMB data allows model-independent testing of large classes of DE/MG models [2512.03314]. Gaussian-process-based pipelines infer $H(z)$, and thereby reconstruct $c(z)$, $\Lambda(z)$, $M_*^2(z)$, mapping directly onto $\alpha$-functions and then to observable signatures in the growth and lensing data.

The most recent constraints from next-generation surveys (DESI, Planck, BOSS, DES, ACT, etc.) indicate broad consistency with $\Lambda$CDM and general relativity—with present bounds $|\alpha_M| \lesssim 0.04$, $|\alpha_B| \lesssim 0.03$, no detectable deviation in $\mu(a,k)$ or $\eta(a,k)$ within uncertainties, and the full suite of perturbation observables favoring canonical gravity [2509.13804].

The EFTofDE further allows for nonparametric reconstruction approaches, where the background and perturbations can be inferred and mapped back onto covariant Lagrangians (quintessence, k-essence, generalized Galileon) using explicit inversion protocols [2507.01442].

## References

- Quantum-Kinetic Dark Energy (QKDE): [2603.14716]
- General EFTofDE formalism and dynamical analysis: [1310.6026][1211.7054]
- Operator basis and $\alpha$-parameterization: [1512.06180][1907.03150]
- Large-scale structure phenomenology and constraints: [1312.6111][1903.08088][2509.13804]
- CEC basis and numerical implementation: [2507.18526]
- Nonlinear regime and N-body codes: [1706.02719][2604.15434]
- Model-independent reconstruction: [2512.03314][2507.01442]

---

The EFTofDE serves as a unifying, stability-prioritized, and model-agnostic approach for testing both dynamical dark energy and modified gravity, providing a framework wherein all viable single-scalar extensions to GR are systematically classified and mapped onto observable cosmological signatures. Its internal consistency, connection to observations, and null-test structure make it an indispensable formalism for exploiting current and upcoming cosmological survey data.

Source: https://www.emergentmind.com/topics/effective-field-theory-of-dark-energy-eftofde