Ghost-Free Mimetic Massive Gravity

• The paper shows that employing a mimetic constraint on the temporal scalar replaces the dangerous sixth mode with mimetic matter, ensuring exactly five propagating massive graviton degrees of freedom.
• It utilizes a non-Fierz–Pauli mass term to avoid the van Dam–Veltman–Zakharov discontinuity and to implement a gravitational Higgs mechanism without ghost instabilities.
• The construction provides a dark-sector interpretation where mimetic matter behaves like dust, offering potential cosmological insights distinct from standard ΛCDM dynamics.

Searching arXiv for the primary paper and closely related follow-up works to ground the article in the cited literature. Ghost-Free Mimetic Massive Gravity is a massive-gravity model in which the graviton mass is generated through a Brout–Englert–Higgs mechanism with four scalar fields, while one of those scalars is constrained as in mimetic gravity. In this construction, the dangerous sixth mode ordinarily associated with the Boulware–Deser ghost is not propagated as an independent scalar ghost; it is constrained and replaced by mimetic matter, so the theory carries only the five degrees of freedom of a massive spin-2 field. A central feature is that the mass term is not of the Fierz–Pauli type, and the model is formulated so that the van Dam–Veltman–Zakharov discontinuity is absent (Chamseddine et al., 2018).

1. Foundational construction

The starting point is the standard “gravitational Higgs” idea for massive gravity. One introduces four scalar fields $\phi^A$ with $A=0,1,2,3$ and forms the diffeomorphism scalar

$$\bar h^{AB}=g^{\mu\nu}\partial_\mu \phi^A \partial_\nu \phi^B-\eta^{AB}.$$

In unitary gauge, $\phi^A=x^A$, the scalar fields are eaten by the metric and provide the longitudinal polarizations of the massive graviton. Around Minkowski space the basic perturbative expansion is

$$g_{\mu\nu}=\eta_{\mu\nu}+h_{\mu\nu}, \qquad \phi^A=x^A+\chi^A,$$

so that the scalar perturbations $\chi^A$ furnish the helicity-0 and helicity-1/Stückelberg components of the massive spin-2 field (Chamseddine et al., 2018).

The distinctive ingredient is the identification of the temporal scalar with a mimetic field, $\phi\equiv \phi^0$, together with the constraint

$g^{\mu\nu}\partial_\mu\phi\,\partial_\nu\phi=1.$

This constraint is implemented by a Lagrange multiplier $\lambda$ in the action

$$I=\int d^4x\,\sqrt{-g}\left[ -\frac12 R+\frac{m^2}{8}\left(\bar h^2-\bar h^{AB}\bar h_{AB}\right) +\lambda\left(g^{\mu\nu}\partial_\mu\phi\,\partial_\nu\phi-1\right) \right].$$

Within the broader mimetic framework, the same condition isolates the conformal mode and produces an effective pressureless-fluid sector; in the massive-gravity setting, it is used as part of the graviton-mass generation mechanism rather than as an additional unconstrained scalar sector (Chamseddine et al., 2018, Malaeb, 15 Feb 2026).

2. Mass term and departure from Fierz–Pauli tuning

In generic massive gravity, the fourth scalar mode is ghostlike unless the quadratic mass term is tuned to the Fierz–Pauli combination

$A=0,1,2,3$0

Even with that tuning, the Boulware–Deser ghost typically reappears nonlinearly. Ghost-Free Mimetic Massive Gravity is built precisely to evade that logic: the dangerous scalar is not removed by Fierz–Pauli tuning, but by the mimetic constraint itself (Chamseddine et al., 2018).

For that reason, the graviton mass term is explicitly non-Fierz–Pauli. The original presentation emphasizes that the effective mass structure

$A=0,1,2,3$1

differs by a sign or relative coefficient from the Fierz–Pauli form, and that this is consistent because the dangerous scalar is not propagating as an independent ghost (Chamseddine et al., 2018). Closely related summaries and follow-up discussions describe the same point by writing altered relative coefficients such as

$A=0,1,2,3$2

or

$A=0,1,2,3$3

while stressing the same structural claim: consistency is achieved through the mimetic constraint rather than through standard Fierz–Pauli tuning (Malaeb, 15 Feb 2026, Chamseddine et al., 2018).

A common misconception is that any healthy massive spin-2 theory must employ the Fierz–Pauli relative coefficient. The mimetic construction is a counterexample in the specific sense claimed by the literature: once the temporal Stückelberg field is constrained mimetically, the usual reason to enforce Fierz–Pauli tuning is absent, because the would-be ghost is no longer an independent propagating mode (Chamseddine et al., 2018).

3. Linearized dynamics and propagating content

The mode analysis is performed around Minkowski space using

$A=0,1,2,3$4

with linearized induced perturbation

$A=0,1,2,3$5

The mimetic constraint fixes the temporal component,

$A=0,1,2,3$6

so the time-like Stückelberg sector is constrained rather than freely propagated (Chamseddine et al., 2018).

The linearized Bianchi identities supply the key relations

$A=0,1,2,3$7

These equations constrain the would-be extra scalar mode. The traceless spatial part

$A=0,1,2,3$8

obeys

$A=0,1,2,3$9

which describes the five polarizations of a massive spin-2 field: two tensor, two vector, and one scalar mode (Chamseddine et al., 2018).

The remaining variables are not independent propagating degrees of freedom. In the original linearized discussion, $$\bar h^{AB}=g^{\mu\nu}\partial_\mu \phi^A \partial_\nu \phi^B-\eta^{AB}.$$0 and $$\bar h^{AB}=g^{\mu\nu}\partial_\mu \phi^A \partial_\nu \phi^B-\eta^{AB}.$$1 are algebraically or constraint-determined in terms of the five physical modes and $$\bar h^{AB}=g^{\mu\nu}\partial_\mu \phi^A \partial_\nu \phi^B-\eta^{AB}.$$2, while the Lagrange multiplier sector satisfies

$$\bar h^{AB}=g^{\mu\nu}\partial_\mu \phi^A \partial_\nu \phi^B-\eta^{AB}.$$3

Accordingly, $$\bar h^{AB}=g^{\mu\nu}\partial_\mu \phi^A \partial_\nu \phi^B-\eta^{AB}.$$4 is interpreted not as a ghost, but as a constrained mimetic-matter component (Chamseddine et al., 2018). The broader review literature presents the same mechanism as the elimination of the dangerous sixth mode, leaving exactly five massive graviton degrees of freedom plus a mimetic dust sector (Malaeb, 15 Feb 2026).

The beyond-linear perturbative treatment reorganizes the same content into scalar, vector, and tensor sectors. In that language, the scalar graviton mode $$\bar h^{AB}=g^{\mu\nu}\partial_\mu \phi^A \partial_\nu \phi^B-\eta^{AB}.$$5 obeys

$$\bar h^{AB}=g^{\mu\nu}\partial_\mu \phi^A \partial_\nu \phi^B-\eta^{AB}.$$6

the vector modes satisfy a massive equation with dispersion relation $$\bar h^{AB}=g^{\mu\nu}\partial_\mu \phi^A \partial_\nu \phi^B-\eta^{AB}.$$7, and the transverse traceless tensor obeys

$$\bar h^{AB}=g^{\mu\nu}\partial_\mu \phi^A \partial_\nu \phi^B-\eta^{AB}.$$8

This formulation again yields five massive graviton polarizations, while the mimetic sector remains separate at linear order (Chamseddine et al., 2018).

4. Ghost removal, Hamiltonian structure, and nonlinear regime

The ghost-free claim is formulated as a nonperturbative statement about the constraint structure. In ordinary massive gravity, the extra sixth scalar mode in $$\bar h^{AB}=g^{\mu\nu}\partial_\mu \phi^A \partial_\nu \phi^B-\eta^{AB}.$$9 becomes the Boulware–Deser ghost at nonlinear order. In the mimetic construction, that mode is not free, because the constraint

$\phi^A=x^A$0

is imposed directly at the level of the action. The role of the sixth mode is transferred to the Lagrange-multiplier sector, which acts as mimetic matter, and this replacement is asserted to persist to all orders in perturbation theory (Chamseddine et al., 2018).

The Hamiltonian analysis makes that logic explicit in canonical language. The linearized theory is formulated in terms of canonical variables for the Stückelberg and metric perturbations, with $\phi^A=x^A$1 as a primary constraint because the Hamiltonian does not depend on $\phi^A=x^A$2. Requiring its conservation yields a secondary constraint, and the total constraint algebra closes under Poisson brackets. The final counting gives five degrees of freedom, matching the expected content of a healthy massive graviton in four dimensions (Malaeb et al., 2019).

The same canonical study emphasizes that $\phi^A=x^A$3 appears as a Lagrange multiplier and contributes additional constraints, and that the Hamiltonian equations reproduce the linearized field equations of the original model. In that presentation, the ghost-free conclusion follows from the chain

$\phi^A=x^A$4

with no extra scalar ghost (Malaeb et al., 2019).

Beyond linear approximation, the model exhibits a nontrivial strong-coupling structure. The detailed perturbative analysis reports that three of the graviton degrees of freedom develop nonlinear corrections comparable to the linear terms already at a length scale of order $\phi^A=x^A$5 in the abstract formulation, while the detailed summary states that the scalar and vector modes become strongly coupled at a scale written as $\phi^A=x^A$6; in both descriptions, the common physical claim is that the three extra polarizations become strongly coupled well before the two transverse tensor polarizations, which remain weakly coupled until the Planck scale (Chamseddine et al., 2018). The same study states that, in the weakly coupled domain, mimetic matter is completely decoupled from the massive graviton and behaves as cold particles of half of the graviton mass (Chamseddine et al., 2018).

5. Absence of the vDVZ discontinuity

A notable property of the model is the absence of the van Dam–Veltman–Zakharov discontinuity. In conventional Fierz–Pauli massive gravity, the massless limit is discontinuous because the extra scalar polarization does not decouple smoothly. In the mimetic theory, the scalar sector is constrained by the Bianchi identities together with the mimetic condition rather than propagated as an unconstrained graviton polarization (Chamseddine et al., 2018).

The relevant relations,

$\phi^A=x^A$7

act analogously to gauge conditions in the linearized analysis. Because the mimetic scalar is not a free propagating polarization of the graviton, the problematic extra coupling responsible for the vDVZ discontinuity in Fierz–Pauli theory is absent already at linear order (Chamseddine et al., 2018). The review literature reiterates that the linearized theory does not suffer from the vDVZ discontinuity in the usual way, precisely because the Bianchi identities and mimetic constraint alter the scalar-sector structure relative to standard massive gravity (Malaeb, 15 Feb 2026).

This feature is conceptually important because it separates the model from the standard massive-spin-2 narrative in two respects at once: the mass term is non-Fierz–Pauli, and the troublesome scalar polarization is not merely tuned away but replaced by constrained mimetic matter (Chamseddine et al., 2018).

6. Cosmological implications and dark-sector interpretation

The construction has an immediate dark-sector interpretation. The original paper notes that, in the linearized regime, the mimetic component behaves like particles at rest with zero momentum and mass equal to half of the graviton mass. This is described as distinct from ordinary cold dark matter and from the dust-like mimetic matter of standard mimetic gravity, while still retaining a dark-sector interpretation (Chamseddine et al., 2018).

The first dedicated cosmological analysis studies a spatially homogeneous and isotropic background with

$\phi^A=x^A$8

where the mimetic constraint enforces $\phi^A=x^A$9 on shell. The background Friedmann equation becomes

$$g_{\mu\nu}=\eta_{\mu\nu}+h_{\mu\nu}, \qquad \phi^A=x^A+\chi^A,$$0

so the mass sector contributes an effective negative cosmological constant, an effective radiation-like term, and an effective curvature-like term, while the integration constant $$g_{\mu\nu}=\eta_{\mu\nu}+h_{\mu\nu}, \qquad \phi^A=x^A+\chi^A,$$1 provides a dust-like mimetic dark matter component (Solomon et al., 2019).

That cosmological analysis also finds a branch structure analogous to other modified gravity models. For $$g_{\mu\nu}=\eta_{\mu\nu}+h_{\mu\nu}, \qquad \phi^A=x^A+\chi^A,$$2, the ghost-free region does not self-accelerate; late-time acceleration must come from a separate positive cosmological constant or another dark-energy sector. For $$g_{\mu\nu}=\eta_{\mu\nu}+h_{\mu\nu}, \qquad \phi^A=x^A+\chi^A,$$3, the model can self-accelerate, but the self-accelerating region is ghostly (Solomon et al., 2019). At the perturbative level, scalar stability requires

$$g_{\mu\nu}=\eta_{\mu\nu}+h_{\mu\nu}, \qquad \phi^A=x^A+\chi^A,$$4

and stability from matter-radiation equality onward gives the bound

$$g_{\mu\nu}=\eta_{\mu\nu}+h_{\mu\nu}, \qquad \phi^A=x^A+\chi^A,$$5

The same study concludes that observational constraints force $$g_{\mu\nu}=\eta_{\mu\nu}+h_{\mu\nu}, \qquad \phi^A=x^A+\chi^A,$$6 and $$g_{\mu\nu}=\eta_{\mu\nu}+h_{\mu\nu}, \qquad \phi^A=x^A+\chi^A,$$7 to be very small, making the model very close to $$g_{\mu\nu}=\eta_{\mu\nu}+h_{\mu\nu}, \qquad \phi^A=x^A+\chi^A,$$8CDM on linear cosmological scales (Solomon et al., 2019).

Within the broader mimetic-gravity literature, ghost-free mimetic massive gravity is therefore treated as a stability-oriented extension. The review literature contrasts it with higher-derivative mimetic modifications such as $$g_{\mu\nu}=\eta_{\mu\nu}+h_{\mu\nu}, \qquad \phi^A=x^A+\chi^A,$$9 or more general $\chi^A$0 models, which can suffer from ghost or gradient instability, and presents the massive-gravity construction as a cleaner route because it does not rely on dangerous higher-derivative operators to stabilize perturbations (Malaeb, 15 Feb 2026).

7. Relation to unimodular and other mimetic extensions

A closely related development is Massive Unimodular Gravity, which is described as a ghost-free massive deformation of unimodular gravity constructed in the spirit of mimetic massive gravity. Its starting point is the unimodular condition

$\chi^A$1

together with Weyl invariance and a mimetic constraint adapted to the unimodular setting. The key distinction is that the no-go theorem for a Lorentz-invariant mass term in unimodular gravity is avoided only by giving up Lorentz invariance (Alvarez et al., 2018).

In that unimodular version, the mimetic degree of freedom does not survive as a propagating on-shell mode. The linearized analysis yields five massive graviton polarizations, no Boulware–Deser ghost, and

$\chi^A$2

so the mimetic sector is absorbed into the constraint structure rather than appearing as a dark-matter-like propagating component (Alvarez et al., 2018). This marks a sharp contrast with standard Ghost-Free Mimetic Massive Gravity, in which the constrained mimetic sector remains physically interpretable as a dust-like component (Chamseddine et al., 2018).

The 2026 review places Ghost-Free Mimetic Massive Gravity alongside mimetic Hořava gravity as one of the major stability-oriented extensions of the mimetic framework. In that comparison, the massive-gravity branch preserves a massive spin-2 description and eliminates the Boulware–Deser ghost through the mimetic constraint plus a non-Fierz–Pauli mass structure, whereas the Hořava-inspired branch uses the mimetic scalar to define a preferred foliation and build covariant higher-spatial-derivative terms (Malaeb, 15 Feb 2026). This suggests that Ghost-Free Mimetic Massive Gravity occupies a distinctive position at the intersection of three themes emphasized repeatedly in the literature: mimetic gravity as a constrained dust sector, Stückelberg or BEH mass generation for the graviton, and nonperturbative control over the dangerous scalar mode through a Lagrange-multiplier constraint rather than through Fierz–Pauli tuning (Chamseddine et al., 2018, Malaeb, 15 Feb 2026).