Existence of ghost-eliminating constraints in multivielbein theory

J. Flinckman; S. F. Hassan

REVIEW 4 major objections 4 minor 25 references

Existence of ghost-eliminating constraints in multivielbein theory

T0 review · 4 major / 4 minor · reviewed 2026-08-04 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A Hamiltonian constraint analysis establishes that multivielbein theory propagates 2+5(N−1) physical modes — one massless and N−1 massive spin-2 fields — with would-be Boulware–Deser ghosts eliminated by a chain of secondary, tertiary, and

desk verdict Careful, honest equal-boost constraint computation with a genuinely new tertiary-constraint result, but the sector is not shown to be dynamically closed and the full ghost-free claim remains deferred. read the letter →

arxiv 2510.03014 v2 pith:N4NNFRDB submitted 2025-10-03 hep-th gr-qc

J. Flinckman , S. F. Hassan This is my paper · ORCID

classification hep-thgr-qc

keywords multivielbeintheoryBoulware–DeserghostHamiltonianconstraintanalysismulti-spin-2bimetricgravitymassivesecond-classconstraintselimination

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper aims to establish that multivielbein theory — N interacting vielbeins with a determinant-type interaction potential — propagates exactly 2+5(N−1) physical degrees of freedom: one massless and N−1 massive spin-2 fields, free of the Boulware–Deser ghosts that typically plague interacting spin-2 theories. The authors perform a Hamiltonian constraint analysis and show that the secondary constraints depend only on dynamical variables, allowing the ghost fields to be eliminated; tertiary constraints, explicitly computed under an equal-boost restriction, eliminate the ghost momenta; and quaternary constraints fix the lapse functions. If correct, the theory would be the first known nonlinear ghost-free theory of N≥3 interacting spin-2 fields beyond pairwise bimetric couplings. The explicit derivation of the tertiary constraints is carried out only in a restricted sector where all boosts coincide, and the paper states that the complete proof beyond this restriction remains to be provided.

What carries the argument

The central mechanism is Dirac's constraint-generation algorithm: starting from primary constraints (vanishing momenta conjugate to the non-dynamical lapse, shift, boost, and rotation fields), one enforces time preservation through Poisson brackets with the total Hamiltonian. The load-bearing structure is the rank-2 matrix M^{IJ} = X^I − X^J, built from trace-kinetic functions X^I of each vielbein's momenta. Because the condition ˙C^I ≈ 0 factors as M^{IJ} N_J ≈ 0, it cannot be solved for the lapses without forcing the interaction potential to vanish; the only consistent solution forces the differences X^I − X^J themselves to vanish, producing N−1 lapse-independent tertiary constraints. The

What would settle it

Compute the Poisson brackets {C^I, C^J_i} and {C^I, J^a_J} without the equal-boost ansatz for N=3. If these fail to vanish weakly, the reduction of ˙C^I ≈ 0 to the rank-2 condition M^{IJ} N_J ≈ 0 breaks down; showing that the quaternary constraints then determine lapses rather than ghost momenta — or, conversely, finding an explicit non-equal-boost background where a sixth ghost mode appears in perturbation theory — would settle the claim.

Watch

Extended reading notes

Core claim

The central discovery is a constraint chain that removes the Boulware–Deser ghost at each of its appearances. The secondary constraints C^I ≈ 0 depend only on the dynamical spatial vielbeins and momenta, not on the lapses, so they can be solved for the ghostly conformal modes. The tertiary constraints C^I_(3)(E,π) ≈ 0 (equation 5.31) are lapse-independent and can be solved for the ghost momenta. The quaternary constraints C^I_(4) ≈ 0 (equation 6.1) then determine N−1 of the N lapses. Counting constraints, the authors find 14 first-class and 22(N−1) second-class constraints, leaving a physical phase space of dimension 2×[2+5(N−1)], which corresponds to one massless and N−1 massive spin-2 fiel

Load-bearing premise

The entire explicit derivation of the tertiary constraints assumes that all boost functions coincide on-shell (p^a_I ≈ p^a), imposed weakly after taking Poisson brackets; if for unequal boosts the equations ˙C^I ≈ 0 instead determine the lapses, the tertiary constraints would not exist and the ghost momenta would propagate.

Editorial extensions

If this is right

If the constraint structure is correct, multivielbein theory (2.19) is a nonlinear theory of one massless and N−1 massive spin-2 fields, free of Boulware–Deser instability for any N.
The mode count 2+5(N−1) matches the quadratic spectrum around proportional backgrounds, so the nonlinear field content is consistent with perturbation theory.
The existence of the tertiary constraints places this theory in the exceptional class (alongside pairwise bimetric couplings) where the would-be lapse-fixing equations instead eliminate ghost momenta; the paper shows this fails in the more general multivielbein models proposed earlier.
The secondary–tertiary–quaternary constraint pattern provides a template for certifying ghost-freedom in other candidate multi-spin-2 theories.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The equal-boost restriction p^a_I ≈ p^a is a symmetry-restricted sector; the generic case is the open question the paper leaves for future work. A natural next step is to recompute the quaternary constraints for N=3 with unequal boosts.
The rank-2 factorization M^{IJ} = X^I − X^J looks like a structural 'miracle' condition: ghost-freedom at the nonlinear level may hinge on just this factorization, suggesting a search for other potentials that share it.
If the assumed simultaneous 3+1 decomposition of all N vielbeins fails for some generic solutions, those configurations would lie outside the current constraint counting, leaving the mode count unproven for them.
One could test the equal-boost sector as a fixed point: if the full constraint algebra shows that unequal-boost configurations are driven toward equal boosts by the dynamics, the sector restriction would be less severe than it appears.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Punchline: the paper does a real, careful Hamiltonian analysis and gets a genuinely new result under the equal-boost Ansatz: N−1 tertiary constraints (5.31) that fix the ghost momenta. But the sector itself is not shown to be dynamically closed, and the general ghost-free claim is explicitly deferred. Read this as a conditional sector result, not a proof of ghost-freedom.

What is good. The computation is disciplined: functional derivatives are taken before imposing p^a_I≈p^a, the rank-2 structure of M^IJ is used honestly, and the exclusion of the lapse-determining branch via eV≠0 is a clean argument. The paper also does a service by presenting the vielbein Hamiltonian formalism in a self-contained way (§3), and the unimodular decomposition in §5.4 makes the ghost variables concrete. The abstract and §7 are honest about the restriction, so no overclaiming there.

The soft spots, in proportion. The big one is the equal-boost sector's consistency. Nothing in the paper shows that the constraint surface p^a_I≈p^a is preserved by the total Hamiltonian. If the boosts separate under evolution, the tertiary constraints derived inside the sector do not constrain the actual dynamical system, and the mode count in §6.3 loses its basis. The paper does not address this. That is load-bearing, not cosmetic. Second, the simultaneous 3+1 decomposition for N>2 is assumed (§2.3), and the paper says proof is future work. Third, the quaternary constraints (6.1), the termination of the algorithm, and the first/second-class classification are asserted with brief argument rather than computed; Table 1 is a summary of claims, not of shown brackets. Fourth, the paper's own conclusion is that the complete proof beyond the Ansatz 'remains to be provided,' so the abstract's first clause should be read with that caveat in mind.

The citation pattern is worth a note but not a charge. The theory under test and the perturbative spectrum check are both from the authors' group. That is not circular, since the Hamiltonian method is independent, but the paper provides no external benchmark for the N≥3 ghost-free expectation.

Who it is for: people working on multi-gravity, ghost-free spin-2 models, and Hamiltonian constraint algorithms. A capable referee who can redo the §5 brackets would be useful; the algebra is long but checkable. I would not desk-reject it. I would send it to review, while being clear to the authors that the dynamical closure of the sector and the N>2 foliation proof are prerequisites for the general claim.

Referee Report

4 major / 4 minor

Summary. The paper performs a Hamiltonian constraint analysis of the multivielbein theory defined by (2.19), with the aim of demonstrating the absence of Boulware–Deser ghosts. It introduces primary constraints for lapses, shifts, boosts and rotations, and derives secondary constraints C^I≈0, C^I_i≈0 and C^I_{\mu\nu}≈0. The authors argue that these constraints determine the rotations and boosts and reduce to lapse-independent constraints that can eliminate the ghost fields. Under an equal-boost Ansatz p^a_I≈p^a, they compute the tertiary constraints ˙C^I≈0 and show, through a rank-2 matrix argument, that consistency requires N−1 lapse-independent constraints C^I_{(3)}=X^I−X^1≈0, which can be solved for the ghost momenta. The stability of these tertiary constraints is asserted to produce quaternary constraints fixing N−1 lapses, leading to the counting 2+5(N−1) physical modes. The paper explicitly assumes a simultaneous 3+1 decomposition for N>2 and concedes that the full proof beyond the equal-boost Ansatz remains future work.

Significance. If the sector result can be promoted to the full theory, this would be the first explicit constraint-level demonstration of ghost elimination in an N≥3 interacting spin-2 theory beyond pairwise bimetric couplings, consistent with the quadratic analysis of [25] and with the original proposal [1]. The explicit computation of the Poisson brackets in Section 5 and the rank-2 kernel argument in Section 5.3 are careful and constitute a nontrivial technical step. The paper is honest about its main limitations, but these limitations are load-bearing: the equal-boost sector is not shown to be dynamically closed, and the simultaneous 3+1 foliation for N>2 is assumed, not proven. Therefore the central claim is conditional, and the full significance depends on closing these gaps.

major comments (4)

[§5.1, §7] The equal-boost Ansatz p^a_I≈p^a is imposed weakly after computing Poisson brackets, but no proof is given that the submanifold p^a_I=p^a is preserved by the Hamiltonian flow. The multipliers λ^a_I are determined by ˙C^I_i≈0 (§4.3), and nothing shown here guarantees they are equal for all I. If they differ, the boosts separate under evolution and the tertiary constraints C^I_{(3)}≈0 (5.31), derived in the sector, do not persist. The §7 concession addresses the generic case, but even the sector result requires dynamical closure. This is load-bearing for the 2+5(N−1) counting in §6.3.
[§2.3] The simultaneous 3+1 decomposition of all N vielbeins is assumed for N>2; the paper notes that it has been established only for N=2 in [10]. The entire Hamiltonian analysis, including the constraint algebra and the phase-space dimension count, depends on this foliation. This assumption should be stated as an explicit hypothesis of the main result, or justified, since without it the analysis may not cover the full solution space.
[§6.1] The quaternary constraints C^I_{(4)} are asserted to be N−1 linear equations for the lapses, but no computation is presented. Since Section 5 derives the tertiary constraints explicitly, the omission of the corresponding ˙C^I_{(3)} computation leaves the termination of the constraint algorithm and the counting in (6.15) on an unverified claim. At least the structure of the bracket {C^I_{(3)}, C^J} should be shown to support the assertion.
[§6.2, Table 1] The classification of the constraints as first-class and second-class is not accompanied by a check of the invertibility of the Dirac matrix of the putative second-class constraints. The paper identifies the diagonal first-class combinations, but the claim that all remaining 22(N−1) constraints are second class requires a nondegeneracy argument. This is needed for the mode count in §6.3.

minor comments (4)

[§2.1] The indices I,J are used for ghost fields in the motivational example and for vielbein species in the rest of the paper. A brief note would avoid confusion.
[§5.4] The statement that the single residual ghost mode is pure gauge and hence non-propagating is plausible but not demonstrated. A short argument or a reference would strengthen this claim.
[Abstract] The abstract mentions the equal-boost restriction but not the simultaneous 3+1 foliation assumption. Since the latter is also load-bearing, it should be disclosed in the abstract or at least in the introduction.
[Eq. (5.22)] The sign convention in the definition of eV should be double-checked. The conclusion that the solved lapses imply V=0 is correct, but the notation could be made clearer to avoid sign ambiguity.

Circularity Check

0 steps flagged · score 1.0 of 10

No circular reduction found: the Hamiltonian constraint derivation is self-contained within the stated equal-boost sector; the self-citations and explicit assumptions are not load-bearing.

full rationale

The central derivation is not circular. Starting from the action (2.19), the paper constructs H_T in (4.8), computes the secondary constraints (4.15)-(4.22), and explicitly evaluates the Poisson brackets needed for stability, obtaining (5.20). The passage from (5.20) to the tertiary constraints (5.31) is a rank-2/kernel argument combined with the non-vanishing-potential assumption eV≠0; it does not define the target counting into the computation. The final mode count (6.15) is a constraint-classification count, not a fit or a renaming. The only external inputs with author overlap are the theory itself, cited to [1], and the N=2 foliation result, cited to [10], but the paper explicitly assumes rather than derives the general-N foliation: "we assume here that it also holds for general N." The quadratic cross-check against [25] is corroboration, not a premise. The paper also honestly flags its main limitation: "a complete proof of the existence and explicit structure of the additional secondary constraints beyond the simplifying Ansatz remains to be provided." This narrows the claim but is not circular. The possible non-closure of the equal-boost sector is a dynamical-consistency question, not a reduction of the prediction to its input. Minor self-citation overlap warrants a low nonzero score, but no circular step can be exhibited; score is therefore 1.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The paper introduces no invented entities and fits no numbers. The load-bearing input structure is: (a) the assumed common foliation for N>2, which the authors flag as an assumption to be proven; (b) the det(u)≠0 invertibility condition, used to reject the alternative solution branch of ˙C^I ≈ 0; and (c) the equal-boost restriction delimiting the sector in which the tertiary constraints are actually computed. The 'ghost' fields are not invented entities: they are the negative-kinetic conformal modes of the spatial vielbeins identified in (5.34).

assumptions (5)

domain assumption A simultaneous 3+1 decomposition exists for all N vielbeins (a common spatial hypersurface), assumed here for N>2.
§2.3: 'we assume here that it also holds for general N'; without it the phase-space decomposition, the lapse/shift split of the potential, and the whole constraint counting do not cover general solutions.
domain assumption u = Σ β_I e_I is invertible throughout phase space, i.e., det(u) ≠ 0.
§2.2 and §5.3: used to exclude the alternative branch of ˙C^I ≈ 0 in which the lapses are determined, because that branch forces the interaction potential V = 2m^4 det(u) to vanish and decouples the vielbeins.
ad hoc to paper Equal-boost Ansatz p^a_I ≈ p^a for all I, imposed weakly only after computing Poisson brackets.
§5.1: the simplifying restriction under which the tertiary constraints (5.31) are explicitly derived; the paper concedes the general case is unproven.
standard math Standard Dirac–Bergmann counting: each first-class constraint removes two phase-space dimensions, each second-class constraint removes one.
§6.3, eq. (6.15): the mode count 2+5(N−1) rests on this counting rule and on the classification asserted in Table 1.
domain assumption A negative-sign kinetic term identifies a ghost mode that must be eliminated.
§5.4: the unimodular decomposition isolates the ghostly pair (φ^I, π^I_φ) via the sign of (π^I_φ)^2 in (5.34); the subsequent counting treats those as the fields to be removed.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Existence of ghost-eliminating constraints in multivielbein theory." pith.science (2026). https://pith.science/paper/N4NNFRDB

@misc{pith2026251003014,
  author       = {Pith},
  title        = {Pith review of: Existence of ghost-eliminating constraints in multivielbein theory},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/N4NNFRDB}},
  note         = {Machine review of arXiv:2510.03014}
}

read the original abstract

We perform a Hamiltonian constraint analysis of the multivielbein theory proposed in [Phys. Rev. Lett. 122 (2019) 251101]. The analysis shows that the secondary constraints have the correct form to eliminate the problematic ghost fields that generally plague theories of interacting spin-2 fields. In particular, the would-be ghost fields are eliminated by the constraints associated with the lapse functions. The subsequent constraints must then determine the ghost momenta and the remaining non-dynamical variables. To establish this, we work with a restricted set of vielbeins with equal boost functions. This allows us to explicitly compute the additional constraints that eliminate the canonical momenta associated with the ghost fields and fix the remaining non-dynamical variables. Our analysis confirms that, subject to this restriction on the boosts, the theory with N interacting vielbeins propagates 2+5(N-1) modes. This corresponds to a nonlinear theory of one massless and N-1 massive spin-2 fields free of Boulware-Deser ghost instabilities.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

25 extracted references · 24 linked inside Pith

[25]

Flinckman and S.F

J. Flinckman and S.F. Hassan,Mass spectrum and linear perturbations of ghost-free multi-spin-2 theory,2410.09439. – 38 –

arXiv
[1]

Hassan and A

S.F. Hassan and A. Schmidt-May,Interactions of multiple spin-2 fields beyond pairwise couplings,Phys. Rev. Lett.122(2019) 251101 [1804.09723]

arXiv 2019
[10]

Hassan and M

S.F. Hassan and M. Kocic,On the local structure of spacetime in ghost-free bimetric theory and massive gravity,JHEP05(2018) 099 [1706.07806]

arXiv 2018
[2]

Boulware and S

D.G. Boulware and S. Deser,Can gravitation have a finite range?,Phys. Rev. D6(1972) 3368

1972
[3]

de Rham and G

C. de Rham and G. Gabadadze,Generalization of the Fierz-Pauli Action,Phys. Rev. D82 (2010) 044020 [1007.0443]

arXiv 2010
[4]

de Rham, G

C. de Rham, G. Gabadadze and A.J. Tolley,Resummation of Massive Gravity,Phys. Rev. Lett.106(2011) 231101 [1011.1232]

arXiv 2011
[5]

Hassan and R.A

S.F. Hassan and R.A. Rosen,On Non-Linear Actions for Massive Gravity,JHEP07(2011) 009 [1103.6055]

arXiv 2011
[6]

Hassan and R.A

S.F. Hassan and R.A. Rosen,Resolving the Ghost Problem in non-Linear Massive Gravity, Phys. Rev. Lett.108(2012) 041101 [1106.3344]

arXiv 2012

Show all 25 references

[7]

Hassan, R.A

S.F. Hassan, R.A. Rosen and A. Schmidt-May,Ghost-free Massive Gravity with a General Reference Metric,JHEP02(2012) 026 [1109.3230]

2012 arXiv
[8]

Hassan and R.A

S.F. Hassan and R.A. Rosen,Confirmation of the Secondary Constraint and Absence of Ghost in Massive Gravity and Bimetric Gravity,JHEP04(2012) 123 [1111.2070]

2012 arXiv
[9]

Hassan and R.A

S.F. Hassan and R.A. Rosen,Bimetric Gravity from Ghost-free Massive Gravity,JHEP02 (2012) 126 [1109.3515]

2012 arXiv
[11]

Hassan and A

S.F. Hassan and A. Lundkvist,Analysis of constraints and their algebra in bimetric theory, JHEP08(2018) 182 [1802.07267]

2018 arXiv
[12]

Khosravi, N

N. Khosravi, N. Rahmanpour, H.R. Sepangi and S. Shahidi,Multi-Metric Gravity via Massive Gravity,Phys. Rev. D85(2012) 024049 [1111.5346]

2012 arXiv
[13]

Schmidt-May and M

A. Schmidt-May and M. von Strauss,Recent developments in bimetric theory,J. Phys. A49 (2016) 183001 [1512.00021]

2016 arXiv
[14]

Baldacchino and A

O. Baldacchino and A. Schmidt-May,Structures in multiple spin-2 interactions,J. Phys. A 50(2017) 175401 [1604.04354]

2017 arXiv
[15]

Niedermann, A

F. Niedermann, A. Padilla and P.M. Saffin,Higher Order Clockwork Gravity,Phys. Rev. D 98(2018) 104014 [1805.03523]

2018 arXiv
[16]

Molaee and A

Z. Molaee and A. Shirzad,Hamiltonian formalism of the ghost free Tri(-Multi)gravity theory, Class. Quant. Grav.38(2021) 065006 [1908.05041]

2021 arXiv
[17]

Dokhani, Z

A. Dokhani, Z. Molaee and A. Shirzad,Gauge generator for bi-gravity and multi-gravity models,Nucl. Phys. B966(2021) 115360 [2001.10947]

2021 arXiv
[18]

Wood, P.M

K. Wood, P.M. Saffin and A. Avgoustidis,Black holes in multimetric gravity,Phys. Rev. D 109(2024) 124006 [2402.17835]

2024 arXiv
[19]

Wood,A new look at multi-gravity and dimensional deconstruction,2501.16442

K. Wood,A new look at multi-gravity and dimensional deconstruction,2501.16442

arXiv
[20]

Hinterbichler and R.A

K. Hinterbichler and R.A. Rosen,Interacting Spin-2 Fields,JHEP07(2012) 047 [1203.5783]. – 37 –

2012 arXiv
[21]

Afshar, E.A

H.R. Afshar, E.A. Bergshoeff and W. Merbis,Interacting spin-2 fields in three dimensions, JHEP01(2015) 040 [1410.6164]

2015 arXiv
[22]

de Rham and A.J

C. de Rham and A.J. Tolley,Vielbein to the rescue? Breaking the symmetric vielbein condition in massive gravity and multigravity,Phys. Rev. D92(2015) 024024 [1505.01450]

2015 arXiv
[23]

Boulanger, S

N. Boulanger, S. Garcia-Saenz, S. Pan and L. Traina,Cubic interactions for massless and partially massless spin-1 and spin-2 fields,JHEP11(2024) 019 [2407.05865]

2024 arXiv
[24]

De Felice, F

A. De Felice, F. Larrouturou, S. Mukohyama and M. Oliosi,Minimal Theory of Bigravity: construction and cosmology,JCAP04(2021) 015 [2012.01073]

2021 arXiv

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.

[1] [25]

Flinckman and S.F

J. Flinckman and S.F. Hassan,Mass spectrum and linear perturbations of ghost-free multi-spin-2 theory,2410.09439. – 38 –

arXiv

[2] [1]

Hassan and A

S.F. Hassan and A. Schmidt-May,Interactions of multiple spin-2 fields beyond pairwise couplings,Phys. Rev. Lett.122(2019) 251101 [1804.09723]

arXiv 2019

[3] [10]

Hassan and M

S.F. Hassan and M. Kocic,On the local structure of spacetime in ghost-free bimetric theory and massive gravity,JHEP05(2018) 099 [1706.07806]

arXiv 2018

[4] [2]

Boulware and S

D.G. Boulware and S. Deser,Can gravitation have a finite range?,Phys. Rev. D6(1972) 3368

1972

[5] [3]

de Rham and G

C. de Rham and G. Gabadadze,Generalization of the Fierz-Pauli Action,Phys. Rev. D82 (2010) 044020 [1007.0443]

arXiv 2010

[6] [4]

de Rham, G

C. de Rham, G. Gabadadze and A.J. Tolley,Resummation of Massive Gravity,Phys. Rev. Lett.106(2011) 231101 [1011.1232]

arXiv 2011

[7] [5]

Hassan and R.A

S.F. Hassan and R.A. Rosen,On Non-Linear Actions for Massive Gravity,JHEP07(2011) 009 [1103.6055]

arXiv 2011

[8] [6]

Hassan and R.A

S.F. Hassan and R.A. Rosen,Resolving the Ghost Problem in non-Linear Massive Gravity, Phys. Rev. Lett.108(2012) 041101 [1106.3344]

arXiv 2012