REVIEW 5 minor 3 references
4D Palatini-Cartan Gravity in Hamiltonian Form
T0 review · 0 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash
Pith's one-line read Under metric nondegeneracy, 4D Palatini-Cartan gravity is classically equivalent to a zero-Hamiltonian constrained system obtained by eliminating an auxiliary connection field.
desk verdict A clean, careful derivation of a known Hamiltonian structure via a genuinely new auxiliary-elimination lemma; not new physics, but solid and useful for the quantization program. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The central mechanism is the change of variables $(e,\omega) \mapsto (\mu,Z,w,e,\omega,v)$ obtained by writing $\hat{\omega} = \omega + v$ with $e v = 0$ and $\epsilon_n d_\omega e = e\sigma$, imported from [CCS21b]: this isolates $v$ as an algebraic, non-propagating field. The quadratic form $S_{\mathrm{aux}} = \int \frac12(\mu\epsilon_n + \iota_Z e)\, e[v,v]$, written pointwise in a gauge where $e^a_\mu = \delta^a_\mu$, becomes the explicit six-variable form $L_{\mathrm{aux}} = xy - xz + yz - 2(\alpha^2+\beta^2+\gamma^2)$, whose nondegeneracy is what makes $v$ eliminable. The symplectic potential $\alpha_\Sigma = \int_\Sigma \frac12 e^2 \delta\omega$ identifies $(e,\omega)$ as the phase-space variables.
What would settle it
Find a metric-nondegenerate coframe $e$ on $\Sigma\times I$ (satisfying $e^3\epsilon_n \neq 0$ and positive-definite $g_\Sigma$) and a nonzero horizontal $(1,2)$-form $v$ with $e v = 0$ such that the pointwise Euler–Lagrange equation derived from $L_{\mathrm{aux}}$ admits a nonzero solution; concretely, at some point the matrix of the quadratic form $L_{\mathrm{aux}} = xy - xz + yz - 2(\alpha^2+\beta^2+\gamma^2)$ would have a zero eigenvalue, or a deformation of the gauge-fixed form would change rank. Any such example would break the claim that $v$ is forced to vanish.
Extended reading notes
Core claim
On a cylinder $\Sigma\times I$, and assuming the coframe $e$ is metric nondegenerate in the sense of Definition 3.2, the paper rewrites the 4D Palatini–Cartan action by decomposing the horizontal connection as $\hat{\omega} = \omega + v$, with the auxiliary part $v$ constrained by $e v = 0$ and the structural constraint $\epsilon_n d_\omega e = e\,\sigma$. The action splits as $S = S_{\mathrm{Ham}}[\mu,Z,w,e,\omega] + S_{\mathrm{aux}}[\mu,Z,e,v]$, where $S_{\mathrm{aux}}$ is quadratic in $v$ with no derivatives. Lemma 3.5 shows that, pointwise and after gauge fixing, the quadratic form is $L_{\mathrm{aux}} = xy - xz + yz - 2(\alpha^2+\beta^2+\gamma^2)$ in six independent components, hence nondegenerate; consequently the Euler–Lagrange equation for $v$ has only the zero solution and $v$ is eliminated. The residual action $S_{\mathrm{Ham}}$ has zero Hamiltonian, with $(e,\omega)$ as the dynamical pair and $(\mu,Z,w)$ as Lagrange multipliers enforcing the torsion-free condition, the Hamiltonian constraint, and the momentum constraint, which are first class. When $\Sigma$ has a boundary, a boundary term appears and reproduces the ADM mass, checked on the Schwarzschild solution and fixing $G_0 = 8\pi G$.
Load-bearing premise
The load-bearing premise is that the metric induced on each slice $\Sigma\times\{t\}$ stays nondegenerate, so that the connection splits uniquely into a torsion-free part and an auxiliary part and the auxiliary quadratic form is nondegenerate; if the metric degenerates, the auxiliary field can no longer be eliminated.
Editorial extensions
If this is right
- Classically, 4D Palatini–Cartan gravity on a cylinder is exactly the constrained Hamiltonian system with zero evolution Hamiltonian; no separate constraint analysis is needed to arrive at it.
- The auxiliary component $v$ of the connection carries no classical degrees of freedom, so the physical phase space is parameterized by $e$ and the reduced connection $\omega$ only.
- The three constraints are first class and generate internal gauge transformations, transversal diffeomorphisms, and spatial diffeomorphisms; there is no time evolution, as expected from diffeomorphism invariance.
- For manifolds with boundary, the boundary term in the modified action is a de facto Hamiltonian; evaluated on Schwarzschild it gives the ADM mass, fixing the constant $G_0 = 8\pi G$.
- In the BV formalism, $v$ can be integrated out first via a BV pushforward, so the subsequent quantization of PC gravity can start from the boundary AKSZ action rather than the full bulk action.
Reading between the lines
- The pointwise nondegeneracy of the auxiliary quadratic form is established in a gauge-fixed frame; a natural extension is to classify all possible normal forms of $L_{\mathrm{aux}}$ for other signatures or dimensions, where the form may change character.
- The paper's derivation suggests that the covariant phase space of 4D gravity can be reconstructed directly from $\alpha_\Sigma$ without a boundary analysis; if so, the symplectic potential would be a canonical starting point for defining quasi-local charges in first-order variables.
- For degenerate metrics, the elimination of $v$ should fail, possibly revealing a sector of the theory where the auxiliary field becomes dynamical; probing that sector could indicate whether the nondegeneracy assumption hides extra degrees of freedom.
- The boundary mass formula $M(r)$ between the horizon and infinity could be compared with other quasi-local mass definitions in the coframe formalism; the comparison is an explicit test of whether this boundary Hamiltonian is the physically correct notion of energy.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper considers 4D Palatini–Cartan gravity on a cylinder M=Σ×I and, under a metric-nondegeneracy condition (Definition 3.2), performs a change of variables based on the decomposition bω=ω+v satisfying the structural constraints (3.5)-(3.6). It rewrites the action as SHam+Saux, shows in Lemma 3.5 that the quadratic form Saux is nondegenerate on the constraint subspace ev=0, and eliminates v to obtain an equivalent Hamiltonian action SHam for the fields (μ,Z,w,e,ω). The resulting theory has zero Hamiltonian and first-class constraints corresponding to the torsion-free condition, the Hamiltonian constraint, and the momentum constraint. With boundaries, the same elimination yields a boundary term that reproduces the ADM mass, and the Schwarzschild check fixes G0=8πG. The paper also briefly discusses Euclidean signature and prospects for quantization via BV pushforward.
Significance. If the central claim holds, the paper gives a transparent and economical derivation of the Hamiltonian formulation of tetrad gravity: the auxiliary field v is eliminated algebraically rather than appearing through a detailed constraint analysis, and the known Hamiltonian and momentum constraints emerge naturally as equations imposed by Lagrange multipliers. The auxiliary-quadratic-form computation in Lemma 3.5 is explicit and checkable, and the boundary analysis correctly recovers the ADM mass, which provides a useful consistency check. The dependence on the decomposition theorem from [CCS21b] is real but is a citation to prior published work rather than an internal gap. The novelty over the authors' earlier papers is modest—this is a pedagogical and technical cleanup of the classical equivalence—but the presentation is careful and the result is useful for subsequent BV/quantization work.
minor comments (5)
- [Section 3, around (3.5)-(3.6)] The existence and uniqueness of the decomposition bω=ω+v satisfying (3.5)-(3.6) is load-bearing for the entire argument, but it is only cited from [CCS21b, Section 4.1]. Please state the theorem explicitly with its precise hypotheses (or include a short proof in an appendix), so that the reader can verify that the change of variables is globally defined and that the structural constraint (3.6a) is indeed part of the parametrization rather than an additional condition to be imposed on SHam.
- [Section 3.1] The text says that v is an auxiliary field and that its Euler-Lagrange equation is a homogeneous linear equation, but after the decomposition theorem v is a determined function of (e,bω), not an independent variable. Please clarify that the auxiliary-field treatment is valid because (3.6) are used to define a coordinate system on the image of the change of variables, and that variations of v are taken along this constrained submanifold; this would prevent a possible misreading that the structural constraint is being dropped.
- [Section 5, item (3)] The statement that metric nondegeneracy is automatic in the Euclidean case is only true if the reference (0,1)-form ϵn is chosen to be the unit normal to the hypersurface determined by e; otherwise e3ϵn can vanish even when gΣ is positive definite. Please specify this choice explicitly.
- [Section 4.1] The dimensional discussion leading to G0=8πG would be clearer if the units (c=1, or c=ħ=1) were stated explicitly; as written, the reader must reconstruct how G0=8πG is dimensionally consistent with the statement that G0 has dimension length squared.
- [Various] There are several small typographical errors: 'known has' should be 'known as' (Section 3.2); 'modifying a theory buy a boundary term' should be 'by' (footnote 14); 'Metric nondegenracy' should be 'Metric nondegeneracy' (Section 5).
Circularity Check
No significant circularity: the elimination of v is an internally proven algebraic result, and the only prior-work dependencies are proved theorems used as independent inputs.
full rationale
The paper's central claim is that, under metric nondegeneracy, 4D Palatini-Cartan gravity is classically equivalent to the Hamiltonian action SHam after elimination of the auxiliary field v. The derivation chain is: decompose e and omega, import a unique-decomposition theorem (bomega = omega + v with epsilon_n d_omega e = e sigma and e v = 0) from [CCS21b], split the action as SHam + Saux (Proposition 3.4), and prove Saux nondegenerate (Lemma 3.5) by an explicit pointwise computation (eq. 3.10). Since Saux is quadratic in v with no derivatives and has a nondegenerate quadratic form on the constrained subspace, the Euler-Lagrange equation forces v = 0; no fitted parameter or hidden use of the target result enters. The cited decomposition theorem is a parameter-free uniqueness statement from prior work by the same group, with stated assumptions (metric nondegeneracy) that do not include the Hamiltonian equivalence being derived; it is therefore independent support, not circularity. The later checks - that constraints (3.12)-(3.14) are first class and that the Schwarzschild boundary term computes the ADM mass - are external benchmarks; the step 'This fixes G0 = 8*pi*G' is a normalization of a dimensionful constant, not a prediction derived from the theory. No equation in the paper reduces by construction to an input or renames a fitted quantity as a prediction. Consequently no circular step is exhibited.
Assumptions & free parameters
assumptions (4)
- domain assumption Unique decomposition bω = ω + v with ϵn dωe = eσ and ev = 0 (structural constraint, (3.5)-(3.6)).
- domain assumption Metric nondegeneracy: e3ϵn ≠ 0 and gΣ positive definite for all t (Definition 3.2).
- domain assumption The constraints (3.12)-(3.14) are first-class and generate gauge, Hamiltonian, and momentum transformations.
- domain assumption αΣ = ∫Σ 1/2 e2 δω is a symplectic potential with symplectic form δαΣ (equation (3.11)).
Cite this review
Pith. "Pith review of 4D Palatini-Cartan Gravity in Hamiltonian Form." pith.science (2026). https://pith.science/paper/22Z4ZIEE
@misc{pith2026250702431,
author = {Pith},
title = {Pith review of: 4D Palatini-Cartan Gravity in Hamiltonian Form},
year = {2026},
howpublished = {\url{https://pith.science/paper/22Z4ZIEE}},
note = {Machine review of arXiv:2507.02431}
}
read the original abstract
In this note the Hamiltonian formulation of four-dimensional gravity, in the Palatini-Cartan formalism, is recovered by elimination of an auxiliary field appearing as part of the connection.
Reference graph
Works this paper leans on
-
[312]
New vari- ables for classical and quantum gravity in all dimensions: I. Hamiltonian analysis
doi: 10.1016/0370-2693(77)90553-6. [BTT13] N. Bodendorfer, T. Thiemann, and A. Thurn. “New vari- ables for classical and quantum gravity in all dimensions: I. Hamiltonian analysis”. Classical and Quantum Grav- ity 30.4 (Jan. 2013), p. 045001. doi: 10 . 1088 / 0264 - 9381/30/4/045001. [CCS21a] G. Canepa, A. S. Cattaneo, and M. Schiavina. “Gen- eral Relativ...
-
[2023]
Phase space for gravity with boundaries
arXiv: 2307 . 04666 [math-ph]. url: https : / / arxiv.org/abs/2307.04666. [CMR18] A. S. Cattaneo, P. Mnev, and N. Reshetikhin. “Per- turbative Quantum Gauge Theories on Manifolds with Boundary”. Communications in Mathematical Physics 357.2 (Jan. 2018), pp. 631–730. doi: 10.1007/s00220- 017-3031-6. [CMW22a] A. S. Cattaneo, P. Mnev, and K. Wernli. “Constrai...
work page Pith review arXiv 2022
-
[2639]
BV Pushforward of Palatini-Cartan gravity
issn: 1424-0661. doi: 10 . 1007 / s00023 - 023 - 01360-8 . url: http://dx.doi.org/10.1007/s00023- 023-01360-8. [CC25] G. Canepa and A. S. Cattaneo. BV Pushforward of Palatini-Cartan gravity. 2025. arXiv: 2507.06279 [math-ph]. url: https://arxiv.org/abs/2507.06279. [CCS21b] G. Canepa, A. S. Cattaneo, and M. Schiavina. “Boundary structure of General Relativ...
work page Pith review arXiv 2021
Reviewed August 6, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.