Rank-Deficit Bias in Gravity Theories

• Rank-deficit bias is the mathematical and physical phenomenon where a constrained kinetic matrix loses full rank, leading to zero eigenvalues and the emergence of ghost modes.
• It is illustrated in massive gravity and quadratic gravity models, where the elimination of degenerate fields induces tachyonic ghost excitations that compromise unitary evolution and vacuum stability.
• The bias significantly impacts cosmological and astrophysical scenarios, as seen in ghost stars and early universe models where negative-energy modes influence observable dynamics.

Rank-deficit bias refers to the set of mathematical and physical pathologies that arise when the kinetic matrix or Hessian associated with a constrained field theory, particularly in gravitational or high-derivative extensions, fails to be of full rank. In the context of gravity and related theories, this bias is manifested through the emergence of ghost degrees of freedom, typically with negative or zero kinetic energy, often resulting from a vanishing determinant of the kinetic operator due to the imposed constraints. Rank-deficit is thus closely tied to the persistence of unphysical (ghost) modes in constrained systems, the failure to decouple dangerous degrees of freedom, and the ensuing breakdown of unitary evolution and vacuum stability.

1. Mathematical Structure and Rank Deficiency in Field Theories

In field theory and gravity, the kinetic matrix (or Hessian) governs the quadratic term in field fluctuations. A rank-deficient Hessian—one with $\det H = 0$—implies the existence of zero modes in the kinetic sector, typically indicating additional symmetries, redundancy, or a constraint structure. However, when physical degrees of freedom are counted incorrectly or constraints are not truly effective, a rank-deficit in the Hessian can signal the re-emergence of ghost-like, negative-norm states.

In "Hidden Ghost in Massive gravity," the authors compute the Hessian $$H_{CD} = \frac{\partial^2 I}{\partial(\partial_0\phi^C)\partial(\partial_0\phi^D)}$$ for a massive gravity action with St\"uckelberg fields, and demonstrate explicitly that $H_{CD}$ possesses a zero eigenvalue since $H_{CD}e^D_0=0$ for a particular linear combination of the fields (where $e^D_0$ is a combination set by the vierbein structure). This results in $\det H=0$ (Chamseddine et al., 2013). The expectation might be that a constrained system with a vanishing determinant would reduce the physical phase space. However, the analysis shows that eliminating the zero-mode actually activates new kinetic terms for the metric which propagate a ghostly excitation.

2. Physical Manifestations: Ghost Excitations from Rank Deficiency

When the Hessian is rank-deficient due to constraints, and a field is algebraically eliminated, extra terms—often with negative kinetic signature—are induced in the remaining action. In the example of massive gravity, the elimination of a degenerate St\"uckelberg scalar generates terms quadratic in $\partial_0 g^{\mu\nu}$, specifically a sign-wrong kinetic term for $h_{00}$, resulting in a tachyonic ghost with mass squared $m_\text{ghost}^2 = -2m^2 < 0$ (Chamseddine et al., 2013).

This process generalizes: any such rank-deficit, if not enforced by a true gauge symmetry or handled carefully, can reintroduce a Boulware–Deser ghost mode. The kinetic matrix’s rank-deficit causes a mismatch between the algebraic counting of degrees of freedom and the actual dynamical propagation, biasing the spectrum toward unphysical, often catastrophic, instabilities.

3. Rank-Deficit Bias in Higher-Derivative and Quadratic Gravity

Quadratic and fourth-order gravity theories typically exhibit rank-deficit bias unless special conditions are met. For the most general action with $R^2$ and $$H_{CD} = \frac{\partial^2 I}{\partial(\partial_0\phi^C)\partial(\partial_0\phi^D)}$$0 terms,

$$H_{CD} = \frac{\partial^2 I}{\partial(\partial_0\phi^C)\partial(\partial_0\phi^D)}$$1

the kinetic matrix contains derivatives up to fourth order. Linearization around flat space and diagonalization of propagator poles reveal that, generically, there is a ghost pole associated with the massive spin-2 sector due to the sign of the residue and the pole structure. The propagator contains $$H_{CD} = \frac{\partial^2 I}{\partial(\partial_0\phi^C)\partial(\partial_0\phi^D)}$$2 terms unless couplings are chosen such that dangerous $$H_{CD} = \frac{\partial^2 I}{\partial(\partial_0\phi^C)\partial(\partial_0\phi^D)}$$3 behavior is removed. In the pure metric (torsionless, "ghostful") theory, no choice of parameters can fully eliminate the ghost without trivializing the theory (Lambiase et al., 20 Oct 2025).

The precise mode content in this context is:

• Two tensor polarizations (massless graviton)
• Scalaron (massive scalar, $$H_{CD} = \frac{\partial^2 I}{\partial(\partial_0\phi^C)\partial(\partial_0\phi^D)}$$4 sector)
• Massive spin-2 ghost (from $$H_{CD} = \frac{\partial^2 I}{\partial(\partial_0\phi^C)\partial(\partial_0\phi^D)}$$5, Ostrogradsky instability)

Rank-deficit bias here ensures that a ghost is always present unless the theory is extended, e.g., to the torsionful case where specific parameter relations can eliminate $$H_{CD} = \frac{\partial^2 I}{\partial(\partial_0\phi^C)\partial(\partial_0\phi^D)}$$6 contributions and yield a unitary, non-ghost spectrum (Lambiase et al., 20 Oct 2025).

4. Rank-Deficit and Residual Ghost Mass in Cosmological Context

A critical consequence of rank-deficit bias in cosmological models is the persistence of relic ghost states after inflation or reheating, with cosmologically disastrous implications. In quadratic gravity with the Weyl ghost, inflation excites a negative-norm ghost that becomes dynamically significant when the Hubble parameter $$H_{CD} = \frac{\partial^2 I}{\partial(\partial_0\phi^C)\partial(\partial_0\phi^D)}$$7 drops below the ghost mass $$H_{CD} = \frac{\partial^2 I}{\partial(\partial_0\phi^C)\partial(\partial_0\phi^D)}$$8. The energy density in the ghost field scales as nonrelativistic matter, redshifting as $$H_{CD} = \frac{\partial^2 I}{\partial(\partial_0\phi^C)\partial(\partial_0\phi^D)}$$9:

$H_{CD}$0

Even for masses in the nominally allowed laboratory window, the ghost energy dominates prior to nucleosynthesis, overclosing the universe. No parameter range exists in which both laboratory and cosmological constraints can be satisfied: rank-deficit bias in the kinetic matrix precludes the decoupling of the ghost (Ivanov et al., 2016).

5. Rank-Deficit Bias in Self-Gravitating Fluid Models ("Ghost Stars")

In relativistic fluid spheres ("ghost stars"), the gravitational mass is given by the Misner–Sharp function

$H_{CD}$1

where the energy–momentum tensor can have regions of negative and positive energy density. The structure of the metric and field equations, under the vanishing complexity condition ($H_{CD}$2), allows for solutions where negative and positive mass precisely cancel at the boundary, producing a total mass $H_{CD}$3 in the limit $H_{CD}$4 (Herrera et al., 4 May 2025, Herrera et al., 2024).

Rank-deficit bias here refers to the inherent tendency of such systems, under imposed constraints (e.g., vanishing total mass, interior sign-changing energy density), to dynamically reach a state where the “observable” mass is strictly zero, but the bulk remains highly nontrivial—manifested by an interior structure that is neither fully trivial nor reducible to Minkowski spacetime, except globally. These models illustrate how internal rank deficits in the stress-energy or metric sector result in boundary-observable quantities (mass, redshift, lensing) vanishing despite nontrivial bulk distribution. No residual ghost mass ("rank deficit remnant") remains once boundary conditions are fully enforced.

6. Observational and Phenomenological Implications

Rank-deficit bias, especially in higher-derivative gravitational theories or ghost star models, leads to distinctive phenomenological signatures or constraints:

• In quadratic gravity, the cancellation between positive-energy and ghost channels nullifies the gravitational wave flux in the massless limit, precluding recovery of the general relativity quadrupole formula. Observational data from pulsar timing and gravitational wave sources thus exclude any viable parameter window for a light ghost (Lambiase et al., 20 Oct 2025).
• In evolving fluid spheres, the approach toward zero total Misner–Sharp mass is accompanied by vanishing surface redshift and disappearance of lensing shadows, providing in principle a clear observational demarcation of the ghost-star end state (Herrera et al., 4 May 2025, Herrera et al., 2024).

7. Summary and Significance

Rank-deficit bias is a foundational pathology in constrained field theories, especially those extending general relativity via additional fields, higher-derivative terms, or tightly imposed algebraic conditions. It is responsible for the dynamical re-introduction of ghost degrees of freedom, both at the level of vacuum structure and in the strong-field, radiative, or thermodynamic evolution of self-gravitating systems. In all known high-energy/cosmological and mathematical implementations, the presence of a persistent rank deficit signals that either unitarity, vacuum stability, or phenomenological viability is compromised, with no consistent parameter window accommodating both experimental and theoretical requirements (Chamseddine et al., 2013, Ivanov et al., 2016, Lambiase et al., 20 Oct 2025, Herrera et al., 4 May 2025, Herrera et al., 2024). The only consistent resolution is either the explicit removal of the offending sector (e.g., via torsion extensions) or the adoption of models where the physical phase-space counting is globally enforced.