Open Gravitational Dynamics

• Open Gravitational Dynamics is a framework treating gravity as an open system that exchanges information through data catalogs, boundary flux, and alternative gauge structures.
• It employs methodologies such as gravitational-wave birefringence tests, Bayesian inference on unbound scattering, and high-precision arithmetic to probe deviations from standard GR.
• Key implications include reformulating gravitational coupling via open EFT, defining canonical brackets through symplectic flux, and integrating quantum memory effects in gravitational interactions.

Searching arXiv for the provided and closely related papers on open gravitational dynamics. {"^^^^^^^^2query2 of Gravitational-Wave Birefringence with the Open Gravitational-Wave Catalog^^^^^^^^2\2 {"^^^^^^^^2query2 gravitational dynamics arXiv","max_results":^^^^2\2query2^^^^} Open gravitational dynamics denotes a family of research programs in which gravitational behavior is treated not as a closed, once-for-all fixed sector of general relativity, but as something that can be inferred, constrained, or reformulated through open data, open-system couplings, boundary flux, unbound scattering observables, or alternative gauge structures. Across the cited literature, the expression covers population-level tests of gravitational-wave propagation with open catalogs, Schwinger–Keldysh and quantum-channel descriptions of gravity in contact with an environment, finite-region gravitational subsystems with nonvanishing symplectic flux, and reformulations in which the gravitational coupling, conformal structure, or computational representation are themselves dynamical (&&&^^^^2query2^^^^&&&, &&&^^^^2\2^^^^&&&, Freidel, 2021, Bini et al., 2020).

^^^^2\2^^^^. Semantic range and unifying motifs

In the cited works, the term appears in several technically distinct senses. One sense is empirical and catalog-based: openly released gravitational-wave data are used to test whether the propagation law of gravitational waves is exactly that of general relativity or instead contains parity-violating birefringent corrections (&&&^^^^2query2^^^^&&&). A second sense is that of open quantum dynamics: a controllable probe PRESERVED_PLACEHOLDER_^^^^2query2^^^^ interacts gravitationally with an inaccessible memory system PRESERVED_PLACEHOLDER_^^^^2\2^^^^, so that the probe experiences a reduced channel

$$\mathcal{E}_t[\rho_P] = \operatorname{tr}_M\!\left[\Xi_t(\rho_P\otimes\rho_M)\right],$$

and the quantumness of gravity is probed through temporal memory effects rather than only through spatial entanglement (Beyer et al., 21 Jul 2025). A third sense is that of dissipative EFT: gravity evolves in the presence of an unobserved environment, and the effective dynamics is formulated on a doubled Schwinger–Keldysh contour with dissipation and noise terms compatible with residual diffeomorphism symmetry (Lau et al., 2024, &&&^^^^2\2^^^^&&&, Christodoulidis et al., 24 Dec 2025).

A separate but related usage concerns finite gravitational subsystems. An open gravitational system can mean a spatial region $\Sigma$ with boundary $S=\partial\Sigma$ through which symplectic flux passes, so that the naïve covariant phase space fails to provide Hamiltonian generators for all corner symmetries. In that setting, openness is literal boundary openness, and the relevant problem is to recover a canonical bracket for charges in the presence of flux (Freidel, 2021).

The literature also uses open more kinematically. In perturbative two-body dynamics, unbound hyperbolic encounters are treated as open conservative systems, and the gauge-invariant scattering angle $\chi(E,J;\nu)$ replaces bound-orbit observables as the primary quantity. This open-system viewpoint is central to sixth-order post-Minkowskian analyses of tail-induced nonlocality in classical scattering (Bini et al., 2020).

A broader structural motif is that gravitational dynamics may itself be open to reformulation. The metric–connection relation need not be fixed a priori, as emphasized in Palatini and metric-affine perspectives, where geodesic and causal structures can be disentangled and the Equivalence Principle becomes a discriminator among competing dynamics (&&&^^^^2\2\2^^^^&&&). Likewise, Shape Dynamics replaces refoliation invariance by spatial conformal invariance, so that the evolving conformal 3-geometry rather than a foliation-independent spacetime becomes the primary dynamical object (&&&^^^^2\22^^^^&&&).

2. Propagation tests, birefringence, and unbound scattering observables

A concrete realization of open gravitational dynamics is the test of gravitational-wave propagation with the 4th-Open Gravitational-wave Catalog. In that analysis, the source dynamics is modeled by standard GR waveforms, but the propagation law is allowed to deviate through parity-violating dimension-5 operators that distinguish left- and right-handed circular polarizations. In circular basis $A=L,R$, the propagation equation is

$$h_A'' + (2+\nu_A)\mathcal{H}h_A' + (1+\mu_A)k^2 h_A = 0,$$

with $\mu_A,\nu_A$ proportional to the helicity label and to the parity-violating scale $M_{\rm PV}$. The observable waveform acquires opposite helicity-dependent phase corrections,

PRESERVED_PLACEHOLDER_^^^^2\2query2^^^^

where PRESERVED_PLACEHOLDER_^^^^2\2\2^^^^ after cosmological propagation (&&&^^^^2query2^^^^&&&).

The data analysis uses exactly 94 events from 4-OGC and Bayesian inference with PyCBC Inference. For the GR baselines, the paper uses IMRPhenomXPHM for BBH, IMRPhenomD^^^^^^^^2\2 for BNS, and IMRPhenomNSBH for NSBH. Excluding the outliers GW^^^^2\2^^^^9^^^^2query2 and GW^^^^2\2^^^^9^^^^2\2\2query2 the combined population constraint is

PRESERVED_PLACEHOLDER_^^^^2\22^^^^

an improvement by a factor of PRESERVED_PLACEHOLDER_^^^^2\2^^^^3 over an earlier twelve-event analysis. GW^^^^2\2^^^^9^^^^2query2 and GW^^^^2\2^^^^9^^^^2\2\2query2 however, show posteriors peaked away from zero, with

PRESERVED_PLACEHOLDER_^^^^2\2^^^^4

for GW^^^^2\2^^^^9^^^^2query2 and

PRESERVED_PLACEHOLDER_^^^^2\2^^^^5

for GW^^^^2\2^^^^9^^^^2\2\2query2 The same analysis finds that assuming birefringent propagation shifts the GW^^^^2\2^^^^9^^^^2query2 sky localization enough that the proposed ZTF flare becomes substantially favored, with

PRESERVED_PLACEHOLDER_^^^^2\2^^^^6

whereas earlier GR-based analyses disfavored the association (&&&^^^^2query2^^^^&&&).

Open kinematics appears again in high-order two-body scattering. Instead of bound orbits, the conservative dynamics is encoded in the hyperbolic scattering angle

PRESERVED_PLACEHOLDER_^^^^2\2^^^^7

treated as a gauge-invariant observable for an unbound encounter. At PRESERVED_PLACEHOLDER_^^^^2\2^^^^8, the relevant nonlocal contribution comes from a hereditary tail Hamiltonian

PRESERVED_PLACEHOLDER_^^^^2\2^^^^9

whose integrated action determines

$$\mathcal{E}_t[\rho_P] = \operatorname{tr}_M\!\left[\Xi_t(\rho_P\otimes\rho_M)\right],$$^^^^2query2^^^^

The analytical evaluation of the 6PM tail contribution requires high-precision arithmetic, PSLQ reconstruction, and direct integration via Harmonic Polylogarithms, and produces transcendental constants up to weight four. This unbound, scattering-based framework is explicitly described as an open-system treatment of conservative binary dynamics, with analytic continuation back to bound EOB observables such as periastron precession (Bini et al., 2020).

3. Gravity as an open quantum process

In the one-sided quantum-memory program, gravity is modeled as a dynamical process between a probe $$\mathcal{E}_t[\rho_P] = \operatorname{tr}_M\!\left[\Xi_t(\rho_P\otimes\rho_M)\right],$$^^^^2\2^^^^ and a memory system $$\mathcal{E}_t[\rho_P] = \operatorname{tr}_M\!\left[\Xi_t(\rho_P\otimes\rho_M)\right],$$2. The total state begins as a product

$$\mathcal{E}_t[\rho_P] = \operatorname{tr}_M\!\left[\Xi_t(\rho_P\otimes\rho_M)\right],$$3

and evolves through a family of CPT maps

$$\mathcal{E}_t[\rho_P] = \operatorname{tr}_M\!\left[\Xi_t(\rho_P\otimes\rho_M)\right],$$4

The reduced two-time dynamics of the probe is

$$\mathcal{E}_t[\rho_P] = \operatorname{tr}_M\!\left[\Xi_t(\rho_P\otimes\rho_M)\right],$$5

The central distinction is between processes realizable with classical memory,

$$\mathcal{E}_t[\rho_P] = \operatorname{tr}_M\!\left[\Xi_t(\rho_P\otimes\rho_M)\right],$$6

and processes that require quantum memory because no such decomposition exists (Beyer et al., 21 Jul 2025).

The witness is one-sided because only the probe is trusted and measured; no assumptions are made about state preparation, measurement, or Hilbert-space dimension on $$\mathcal{E}_t[\rho_P] = \operatorname{tr}_M\!\left[\Xi_t(\rho_P\otimes\rho_M)\right],$$7. Using the Choi operators $$\mathcal{E}_t[\rho_P] = \operatorname{tr}_M\!\left[\Xi_t(\rho_P\otimes\rho_M)\right],$$8, the paper constructs a linear functional

$$\mathcal{E}_t[\rho_P] = \operatorname{tr}_M\!\left[\Xi_t(\rho_P\otimes\rho_M)\right],$$9

which can be written operationally as

$\Sigma$^^^^2query2^^^^

If $\Sigma$^^^^2\2^^^^, then no classical-memory realization exists, and the reduced gravitational dynamics necessarily involves quantum memory (Beyer et al., 21 Jul 2025).

The illustrative gravitational setup consists of two spatial qubits interacting through the Newtonian Hamiltonian

$\Sigma$2

with

$\Sigma$3

A bare gravitational interaction generates only reduced dephasing and remains classically simulable. The nonclassical regime appears only after interleaving local phase gates $\Sigma$4 and a local $\Sigma$5 gate, producing a two-time process whose witness evaluates to

$\Sigma$6

At the ideal point $\Sigma$7, one gets

$\Sigma$8

which certifies that gravity has coherently stored and returned quantum information. This witness probes temporal quantum correlations and is explicitly stated to cover a quantum signature not fully captured by entanglement-based Bose–Marletto–Vedral-type proposals (Beyer et al., 21 Jul 2025).

4. Schwinger–Keldysh, dissipation, and open gravitational EFT

A second major line of work formulates open gravitational dynamics directly in the Schwinger–Keldysh formalism. The guiding assumption is that the system sector—typically gravity plus a single clock or a dissipative scalar—is coupled to an unobserved environment whose microscopic dynamics is integrated out. In the Keldysh basis, the metric is doubled into a retarded field $\Sigma$9 and an advanced field $S=\partial\Sigma$^^^^2query2^^^^, and the effective action takes the schematic form

$S=\partial\Sigma$^^^^2\2^^^^

The linear term defines the effective equations of motion, while the quadratic term defines the noise kernel. Because gravity couples universally, the environment stress tensor must be included explicitly; otherwise a dissipative scalar or dissipative tensor sector becomes inconsistent once full dynamical gravity is restored (Lau et al., 2024, &&&^^^^2\2^^^^&&&).

The gravitational open EFT of inflation sharpens this point. Although many open operators are allowed by residual time-dependent spatial diffeomorphisms, most of them overconstrain the scalar sector and eliminate the propagating curvature mode. The consistent minimal operator is the trace-adjusted extrinsic-curvature combination

$S=\partial\Sigma$2

which satisfies the exact geometric identity

$S=\partial\Sigma$3

This leads to the deformed diffeomorphism identity

$S=\partial\Sigma$4

which preserves the correct scalar and tensor degree count. The resulting scalar perturbation equation becomes a Langevin-type equation for the curvature mode with explicit dissipation and noise (Christodoulidis et al., 24 Dec 2025).

In linear scalar perturbations, the minimal open model yields

$S=\partial\Sigma$5

and

$S=\partial\Sigma$6

This is the fully gravitational extension of dissipative inflationary EFTs that were previously derived only in the decoupling limit (Christodoulidis et al., 24 Dec 2025).

When the environment is modeled hydrodynamically, the same SK machinery gives a unified dissipative EFT for scalar fields, gravitational waves, and black holes. The Einstein equation takes the form

$S=\partial\Sigma$7

while the environment obeys

$S=\partial\Sigma$8

For tensor modes, the general open-system propagation equation derived around cosmological backgrounds is

$S=\partial\Sigma$9

with modified speed $\chi(E,J;\nu)$^^^^2query2^^^^, dissipation $\chi(E,J;\nu)$^^^^2\2^^^^, tensor noise $\chi(E,J;\nu)$2, and parity-violating coefficient $\chi(E,J;\nu)$3. A central result is that, unlike in electromagnetism, the leading gravitational birefringence is dissipative, whereas conservative birefringence appears only at higher derivative order (Lau et al., 2024, &&&^^^^2\2^^^^&&&).

5. Finite regions, symplectic flux, and canonical structure

In the covariant phase-space literature, an open gravitational system is a finite spatial region $\chi(E,J;\nu)$4 whose boundary $\chi(E,J;\nu)$5 is not isolated. The basic structures are the Lagrangian $\chi(E,J;\nu)$6-form $\chi(E,J;\nu)$7, the symplectic potential current $\chi(E,J;\nu)$8, and the symplectic current

$\chi(E,J;\nu)$9

For a diffeomorphism generator $A=L,R$^^^^2query2^^^^, the Noether charge on $A=L,R$^^^^2\2^^^^ is

$A=L,R$2

and the fundamental integrated relation is

$A=L,R$3

where the symplectic flux is

$A=L,R$4

When $A=L,R$5, the region is open and the charge $A=L,R$6 is not Hamiltonian on the naïve bulk phase space (Freidel, 2021).

The resolution is to enlarge the phase space by an embedding field

$A=L,R$7

and its Maurer–Cartan field-space connection

$A=L,R$8

The Lagrangian is extended to

$A=L,R$9

and the dressed symplectic potential becomes

$$h_A'' + (2+\nu_A)\mathcal{H}h_A' + (1+\mu_A)k^2 h_A = 0,$$^^^^2query2^^^^

The corresponding extended symplectic current satisfies

$$h_A'' + (2+\nu_A)\mathcal{H}h_A' + (1+\mu_A)k^2 h_A = 0,$$^^^^2\2^^^^

so that the canonical relation is promoted to

$$h_A'' + (2+\nu_A)\mathcal{H}h_A' + (1+\mu_A)k^2 h_A = 0,$$2

with no flux term on the right-hand side (Freidel, 2021).

This construction yields a genuine canonical bracket for all extended corner symmetries, including normal supertranslations. The extended corner symmetry group is

$$h_A'' + (2+\nu_A)\mathcal{H}h_A' + (1+\mu_A)k^2 h_A = 0,$$3

and the canonical Poisson bracket on the extended phase space,

$$h_A'' + (2+\nu_A)\mathcal{H}h_A' + (1+\mu_A)k^2 h_A = 0,$$4

is shown to coincide with the previously introduced generalized Barnich–Troessaert bracket. In this sense, the openness associated with gravitational radiation and boundary flux is reabsorbed into embedding degrees of freedom, restoring a Hamiltonian description for the full corner symmetry algebra (Freidel, 2021).

6. Alternative realizations, reformulations, and infrastructures

Open gravitational dynamics also appears in string-theoretic settings. For classical open strings ending on D3-branes in a Poincaré $$h_A'' + (2+\nu_A)\mathcal{H}h_A' + (1+\mu_A)k^2 h_A = 0,$$5 throat, the D3-branes’ backreaction makes the proper string length infinite while keeping the inertial mass finite. The resulting gravitational interaction between parallel open strings is dominated by the near-horizon tail and behaves at large separation as

$$h_A'' + (2+\nu_A)\mathcal{H}h_A' + (1+\mu_A)k^2 h_A = 0,$$6

independent of the strings’ inertial masses in non-compact space. After compactification that produces an effective four-dimensional graviton zero mode, the potential becomes

$$h_A'' + (2+\nu_A)\mathcal{H}h_A' + (1+\mu_A)k^2 h_A = 0,$$7

so the long open-string tail induces an equivalence-principle-violating correction to Newtonian gravity. The effect is explicitly identified as specific to string theory and absent for point particles (Benichou, 2010).

A different nonlocal realization replaces Newton’s constant by a dynamical coupling $$h_A'' + (2+\nu_A)\mathcal{H}h_A' + (1+\mu_A)k^2 h_A = 0,$$8 determined by the geometry of parametrized null geodesics through $$h_A'' + (2+\nu_A)\mathcal{H}h_A' + (1+\mu_A)k^2 h_A = 0,$$9. In the full definition,

$\mu_A,\nu_A$^^^^2query2^^^^

and the modified Einstein equation is

$\mu_A,\nu_A$^^^^2\2^^^^

with a trace-free DGC tensor $\mu_A,\nu_A$2 constructed from line integrals along null geodesics. The resulting Einstein–matter–DGC system is shown to be locally well-posed as a quasilinear symmetric hyperbolic system. In FRW symmetry, the simplified model gives

$\mu_A,\nu_A$3

while in the solar system the relative correction is of order $\mu_A,\nu_A$4, hence too small for present-day experiments (Finster et al., 2016).

The same broad field includes structural reformulations of gravitational degrees of freedom. In Palatini and metric-affine approaches, spacetime is described by the triple $\mu_A,\nu_A$5, and the connection is not assumed a priori to be Levi–Civita. This disentangles geodesic structure from causal structure and places the Equivalence Principle at the center of experimental discrimination among theories, with explicit attention to possible quantum-level violations (&&&^^^^2\2\2^^^^&&&). Shape Dynamics pushes the reformulation further by replacing GR’s relativity of simultaneity with invariance under spatial conformal transformations. Through the Linking Theory, ADM GR and Shape Dynamics are obtained as different gauge fixings of the same enlarged phase space; the resulting Shape Dynamics theory keeps the momentum constraints, introduces local conformal constraints, and replaces the local Hamiltonian constraints by a single global Hamiltonian (&&&^^^^2\22^^^^&&&).

The open character of the field is also methodological. Gravitas is an open-source computational GR framework in the Wolfram Language designed to move seamlessly between symbolic and numerical representations of arbitrary spacetime geometries, matter fields, and evolution equations. It implements tensorial objects such as MetricTensor, ChristoffelSymbols, RiemannTensor, RicciTensor, EinsteinTensor, and StressEnergyTensor, solves vacuum and matter-coupled Einstein equations symbolically and numerically, and is explicitly designed to interface with ADM decomposition and a totally unstructured adaptive refinement scheme based on hypergraph rewriting. This provides an open computational substrate for exploring gravitational dynamics without committing to a single fixed coordinate system, topology, or workflow (Gorard, 2023).

Taken together, these strands show that open gravitational dynamics is not a single theory but a technically coherent family of programs. What unifies them is the decision to treat gravitational propagation, interaction, or phase-space structure as something that can exchange information or flux with an environment, boundary, hidden sector, or enlarged gauge structure, rather than as a closed metric system fixed once and for all by standard GR.