Classification and Reduction of Homogeneous Star Products

Chiara Esposito; Jonas Schnitzer; Marvin Dippell

REVIEW 2 major objections 5 minor 8 references

Classification and Reduction of Homogeneous Star Products

T0 review · 2 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read The paper establishes that homogeneous star products on the dual of a Lie algebroid are classified, up to homogeneous equivalence, by a single class in second Lie algebroid cohomology, and that the projectable ones make quantization…

desk verdict Solid homogeneous classification on Lie algebroids; the projectable half needs a real proof where the text invokes a phantom 'Proof XXIII'. read the letter →

arxiv 2507.02820 v1 pith:OKX7MWVU submitted 2025-07-03 math.QA math-phmath.DGmath.MP

Marvin Dippell , Chiara Esposito , Jonas Schnitzer This is my paper · ORCID

classification math.QAmath-phmath.DGmath.MP MSC 53D5553D1717B6616E40

keywords deformationquantizationstarproductshomogeneousLiealgebroidscharacteristicclassescoisotropicreductioncommuteswithHochschild-Kostant-Rosenbergtheorem

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

On the dual $A^*$ of a Lie algebroid $A$, the linear Poisson bracket is the classical side of many deformation quantizations, i.e. associative products on $C^\infty(A^*)[[\hbar]]$ deforming the pointwise product. This paper's central claim is that the homogeneous ones — those respecting the fibrewise dilation of $A^*$ — are completely classified by one characteristic class in $H^2(A)$: two homogeneous star products are equivalent if and only if their classes agree, and every class is realised. The paper then treats projectable homogeneous star products, quantizations that survive a coisotropic reduction to a reduced phase space, and proves that they are classified by the projectable part of $H^2$, in a way that makes reduction commute with quantization. If correct, this turns a question about infinite operator series into a finite cohomological check and identifies exactly when a quantization of a constrained system descends to the reduced space.

What carries the argument

The engine is a homogeneous version of the Hochschild–Kostant–Rosenberg deformation retract: for a torsion-free homogeneous covariant derivative $\nabla$ on $A^*$, operators $H_\nabla$ and $\mathrm{hkr}_\nabla^{-1}$ satisfy the homotopy equation $\partial H_\nabla+H_\nabla\partial=\mathrm{id}-\mathrm{hkr}\circ\mathrm{hkr}_\nabla^{-1}$ and preserve homogeneity degrees, so any closed Hochschild cocycle with vanishing antisymmetrization is exact in the homogeneous category. The characteristic class $\Phi$ compares a star product with the reference product $\star^{(\nabla,0)}$ built in Proposition 3.10, so the class is independent of the chosen connection. For the reduction statement the new object is a constraint Lie algebroid $A_T\leftarrow A_N\twoheadrightarrow A_{\mathrm{red}}$ — a subalgebroid $A_N$ over a submanifold mapping onto a quotient algebroid — and the projectable subcomplex $\Gamma^\bullet_{\mathrm{proj}}(A_T^*)$ of Lie algebroid forms that are pulled back from $A_{\mathrm{red}}$; its cohomology $H^2_{\mathrm{proj}}(A_T)$ is the classifying space for projectable classes.

What would settle it

Run the omitted extension step on the Hopf-fibration constraint Lie algebroid from Example 4.15: if for some closed projectable two-form $B$ the representation of $\star^{(\nabla,B)}\otimes\star_{\mathrm{red}}^{\mathrm{opp}}$ cannot be extended from the generators of the vanishing ideal of $A^\vee_N$ to all functions, then Theorem 4.19's surjectivity claim fails; similarly, two homogeneous star products on some $A^*$ with equal characteristic class but no homogeneous equivalence would refute Theorem 3.17.

Watch

Extended reading notes

Core claim

The central discovery is a bijection $\Phi:\mathrm{Def}_{\mathrm{hom}}(A^*)\to H^2(A)$, where $\mathrm{Def}_{\mathrm{hom}}(A^*)$ is the set of homogeneous equivalence classes of homogeneous star products deforming the linear Poisson structure of $A$. The class is read off from the skew-symmetric part of the second-order bidifferential operator via $\Phi(\star)(s,t)=\iota^*C_2^-(J(s),J(t))$, and Theorem 3.17 states that two homogeneous star products are homogeneously equivalent exactly when these classes coincide, with every cohomology class realised. A direct corollary is that ordinary equivalence of homogeneous star products already forces homogeneous equivalence. For a constraint Lie algebroid $A_T\leftarrow A_N\twoheadrightarrow A_{\mathrm{red}}$, Theorem 4.19 upgrades the bijection to $\Phi^{\star_0}_{\mathrm{proj}}:\mathrm{Def}_{\mathrm{proj}}(A_T^*)\to H^2_{\mathrm{proj}}(A_T)$, with commuting squares that make quantization commute with reduction. The paper also exhibits two projectable star products that are equivalent as homogeneous star products but have inequivalent reductions, which is why projectable equivalences are the right notion.

Load-bearing premise

Everything rests on the imported homogeneous Hochschild–Kostant–Rosenberg deformation retract preserving homogeneity degrees; if that retract does not exist as stated, the exactness arguments collapse, and the projectable classification additionally depends on an omitted step that extends a vanishing condition from generators of the vanishing ideal to all functions.

Editorial extensions

If this is right

Every homogeneous deformation of a linear Poisson structure on $A^*$ is determined, up to homogeneous equivalence, by one element of $H^2(A)$, so comparing two quantizations reduces to a finite cohomology calculation.
A homogeneous star product is homogeneously equivalent to a projectable one exactly when its characteristic class has a projectable representative, i.e. when $I^*\Phi(\star)=P^*[B_{\mathrm{red}}]$; this makes reducibility a cohomological condition.
For constraint Lie algebroids, quantization commutes with reduction: reducing the quantized class and then quantizing gives the same class as quantizing and then reducing, in the projectable classification.
Ordinary equivalence already implies homogeneous equivalence for homogeneous star products, so the refined classification does not distinguish more than the coarse one.
The polynomial functions form a subalgebra of every homogeneous star product, so the formal deformation parameter can be evaluated at complex values, which connects the classification to convergent deformation quantization.

Reading between the lines

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

Editorial inference: the explicit formula $\Phi(\star)(s,t)=\iota^*C_2^-(J(s),J(t))$ offers a practical algorithm for computing the invariant of any concrete star product on the dual of a Lie algebroid, so known constructions could be compared entirely from their second-order data.
Editorial inference: because the proof of projectable surjectivity relies on an omitted extension step, a natural test is whether the classification holds under weaker regularity assumptions on the constraint Lie algebroid; the theorem as stated would survive if that step is supplied.
Editorial inference: the same cohomological mechanism should apply to other settings where a homogeneity-preserving HKR retract exists, for example graded or super versions of Lie algebroids, since the order-by-order arguments use only that retract.
Editorial inference: the projectability criterion yields a possible new obstruction — if the characteristic class of a homogeneous star product is not in the image of $H^2_{\mathrm{proj}}(A_T)\to H^2(A_T)$, no homogeneous equivalence can make it compatible with reduction — which would give a linear analogue of the known examples where reducible quantizations do not exist.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

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

Read the Dippell–Esposito–Schnitzer paper. The main news is the classification of homogeneous star products on duals of Lie algebroids by H^2(A), plus a projectable version by H^2_proj that would make quantization commute with reduction. Section 3 is the real contribution: the construction of star products from a closed B, the relative class, and the bijection in Theorem 3.17 are carefully done. This genuinely extends earlier symplectic and cotangent-bundle classifications. I also like Section 3.3, where they compute characteristic classes for the Neumaier–Waldmann, Kontsevich, and Gutt constructions; Example 3.22 honestly flags that the enveloping-algebra route isn't known to produce bidifferential operators except in the Lie algebra case.

The soft spot is exactly where the stress test points. In Theorem 4.19, surjectivity of the projectable characteristic class map requires showing that a certain obstruction vanishes on the whole vanishing ideal of A_N^vee. The manuscript calculates the obstruction on generators of the form pr^*F - pr^*Ftilde, then says "Proof XXIII is enough" to pass to all functions. No such proof appears. I agree the claim is plausible: the obstruction sits in Lambda^2 nu and should only see conormal differentials. But as written it's an unproved assertion at a load-bearing point. If it fails, the projectable classification and the reduction-commutation diagram are not established. The authors need to supply a real argument there, or state the local computation and prove it.

The other caveat is the heavy reliance on their own preprint [Dip+24] for the homogeneous and representable HKR deformation retracts. That's not circular—those theorems are independent of the classification—but it does mean the paper is not self-contained. The referee should have [Dip+24] in hand. A couple of "one can check" steps in Section 4.1 are also terse, but they look routine.

Verdict: this deserves a serious referee. The core classification is solid and useful, and the reduction-commutation conjecture is worth settling. With the missing proof supplied, I'd accept. I'd tell the editor to send it to someone who can verify the coisotropic reduction step carefully.

Referee Report

2 major / 5 minor

Summary. The paper presents a classification of homogeneous star products on the total space of the dual of a Lie algebroid, showing that homogeneous equivalence classes are in bijection with the second Lie algebroid cohomology. It then introduces projectable star products for constraint Lie algebroids and proves a parallel classification in terms of the projectable cohomology subcomplex, with a commuting diagram that is interpreted as quantization commuting with reduction. The proofs rely heavily on a homogeneous Hochschild-Kostant-Rosenberg deformation retract imported from the authors' preprint [Dip+24], and the final surjectivity step of the projectable classification invokes an unspecified 'Proof XXIII'.

Significance. If the results are correct, they provide a clean cohomological parametrization of homogeneous deformation quantizations of linear Poisson structures and a cohomological criterion for projectability, which would be a substantial advance over the known symplectic and cotangent-bundle cases. The paper is honest about its limitations: Example 3.22 explicitly states that the universal-enveloping-algebra construction is not yet known to be bidifferential, and the dependence on [Dip+24] is openly declared. The treatment of examples, including the Neumaier-Waldmann and Kontsevich-Dolgushev constructions, adds useful context. However, the missing 'Proof XXIII' at a load-bearing point and the heavy reliance on an external preprint prevent the paper from being fully verifiable in its present form.

major comments (2)

[Section 4.2, proof of Theorem 4.19] The surjectivity proof contains the sentence 'Proof XXIII is enough to show that (I∨ × P∨)∗(Cα,red2)− vanishes on the whole vanishing ideal.' No such proof appears in the manuscript. This is load-bearing: the preceding computation only verifies vanishing on the functions pr∗A∗T(F) − pr∗A∗red(F˜), which are asserted to generate the vanishing ideal, while Proposition 4.5 requires the obstruction hkr−1∇(D2) to vanish on all pairs of functions in that ideal. Since this step is exactly what converts representability of ⋆α ⊗ ⋆oppred into the projectable classification and the commuting diagram (4.29), Theorem 4.19 and the claim that quantization commutes with reduction are not established as written. Please supply the missing argument or a precise reference.
[Sections 2.1, 2.2 and Appendix B] The proofs throughout depend on deformation retract data (Theorem 2.3, Proposition 2.6, and Theorem B.1) imported from the authors' preprint [Dip+24]. These results supply the operators H∇ and hkr−1∇ whose homogeneity preservation is used in every exactness argument, for example in Proposition 3.9, Proposition 3.11, Lemma 3.14, Proposition 4.5, and the representable HKR theorem in Appendix B. Because [Dip+24] is not included in this manuscript and appears to be an unreviewed preprint, the homogeneous classification and the projectable classification are conditional on external results that are neither proved nor fully stated here. Please state the precise theorems used and either prove them or document their accepted publication status.

minor comments (5)

[Remark 3.19(ii)] 'homogeneous tar products' should be 'homogeneous star products'.
[Definition 4.9] The heading 'Projectable equivalance' contains a typo; it should be 'Projectable equivalence'.
[Proof of Theorem 4.19] The sentence 'we may assume that I∗dα = 0, since Φ(⋆red) = [Bred]' is terse; a brief explanation of why the chosen α can be adjusted to satisfy this condition would improve clarity.
[Proposition 4.16(ii)] The assertion that projectability of the equivalence implies α ∈ Γ∞proj(A∗T) is stated without proof; a short argument or a reference to an equation would be helpful.
[Example 4.15] It would be clearer to specify which representative of the Fubini-Study class [ω] and which pullback to S2n+1 are used when constructing the two projectable star products.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular reduction: the homogeneous classification is constructed from the HKR retract rather than assumed; the main caveats are a non-circular dependency on the authors' [Dip+24] and an omitted 'Proof XXIII' in Theorem 4.19.

full rationale

The claimed derivation chain is not circular. Section 3 builds the classification from explicit ingredients: the HKR deformation retract (Thm. 2.3, imported from [Dip+24]) yields the homotopy operators; Prop. 3.10 constructs a homogeneous star product star(∇,B) for every closed B; Cor. 3.16 computes its class; Lemma 3.14 proves equivalence iff the relative class is zero; Thm. 3.17 assembles these into the bijection Φ: Def^hom(A*) → H^2(A). The characteristic class is not defined as 'the cohomology class of the product' and then asserted to classify; it is defined via a reference product star(∇,0), shown to be well-defined, and the inverse direction is the explicit construction. No parameter is fitted and no 'prediction' is read back from data used to build the product. The heavy use of [Dip+24] is a self-citation, but it is not a circular reduction: the cited deformation retract is a parameter-free statement about differential Hochschild cohomology with assumptions (manifold, torsion-free connection, linear submanifold in App. B) that do not include the classification of homogeneous star products. Per the reviewing rule, the load-bearing omission is flagged: in the surjectivity proof of Thm. 4.19 (p. 28), the manuscript checks the vanishing of (I^∨ × P^∨)∗(C^{α,red}_2)^- on the functions pr*_{A*_T}(F)−pr*_{A*_red}(F̃) generating the vanishing ideal, then states 'Proof XXIII is enough to show that (I^∨ × P^∨)∗(C^{α,red}_2)^- vanishes on the whole vanishing ideal'; no such proof appears anywhere in the paper. This is a correctness gap, not a circularity: extending from generators to the whole ideal is a lemma about sections of Λ²ν, and the conclusion is not presupposed. Likewise, the 'quantization commutes with reduction' statement (Rem. 4.20) is an interpretation of the commuting diagram (4.29), not an input to the proof. Score 2 reflects the non-circular but real self-citation/dependency and the omitted proof; it is not a claim that a derivation reduces to its own inputs.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

No fitted numerical parameters appear. The inputs such as homogeneous covariant derivatives, closed two-forms B, and the reference star product star_0 are structural choices whose independence is proved in the paper. The load-bearing assumptions are the imported HKR deformation retracts and the constraint Lie algebroid framework, rather than ad hoc axioms invented for this result. No new particles, forces, or unexplained entities are introduced.

assumptions (4)

domain assumption Homogeneous HKR deformation retract from [Dip+24], Theorem 2.3 and Proposition 2.6: there exist homogeneous homotopy operators yielding a deformation retract of the differential Hochschild complex onto multivector fields.
Used throughout Section 3 for exactness of closed bidifferential operators and for vanishing of lower-degree homogeneous parts; imported without proof in this paper.
domain assumption Representable HKR deformation retract from [Dip+24], Appendix B, Theorem B.1: the Hochschild complex with coefficients in differential operators on a linear submanifold deformation-retracts to sections of the normal bundle.
Used in Proposition 4.5 and Proposition 4.11 to extend representations of star products and to solve exactness conditions in the projectable setting.
domain assumption Constraint Lie algebroids and the identification of compatible constraint vector bundles with duals of constraint Lie algebroids, from [DK25].
Throughout Section 4.2 the paper assumes this framework to define projectable forms and H^2_proj(AT); the correspondence is stated as known rather than reproved.
standard math Standard Hochschild-Kostant-Rosenberg theorem computing Hochschild cohomology of smooth functions by multivector fields.
Assumed at the start of Section 2.1 as background; the improved version from [Dip+24] is the main tool built on it.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Classification and Reduction of Homogeneous Star Products." pith.science (2026). https://pith.science/paper/OKX7MWVU

@misc{pith2026250702820,
  author       = {Pith},
  title        = {Pith review of: Classification and Reduction of Homogeneous Star Products},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OKX7MWVU}},
  note         = {Machine review of arXiv:2507.02820}
}

read the original abstract

We present a classification of homogeneous star products on duals of Lie algebroids in terms of the second Lie algebroid cohomology. Moreover, we extend this classification to projectable star products, i.e., to quantizations compatible with (coisotropic) reduction. This implies that quantization commutes with reduction in the considered setting.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

8 extracted references · 8 canonical work pages

[1]

Quantum Me- chanics as a Deformation of Classical Mechanics

[Bay+77] Bayen, F., Flato, M., Frønsdal, C., Lichnerowicz, A., and Sternheimer, D. “Quantum Me- chanics as a Deformation of Classical Mechanics”. In:Lett. Math. Phys.1 (1977), pp. 521–

work page 1977
[5]

Convergence of the Gutt Star Product

[ESW17] Esposito, C., Stapor, P., and Waldmann, S. “Convergence of the Gutt Star Product”. In:J. Lie Theory27 (2017), pp. 579–622. [ESW19] Esposito, C., Schmitt, P., and Waldmann, S. “Comparison and Continuity of Wick-type Star Products on Certain Coadjoint Orbits”. In:Forum Math.31.5 (2019), pp. 1203–1223. [Fed94] Fedosov, B. V. “A Simple Geometrical Con...

work page 2017
[74]

Characterstic classes of star products on Marsden-Weinstein reduced symplectic manifolds

[Rei17] Reichert, Thorsten. “Characterstic classes of star products on Marsden-Weinstein reduced symplectic manifolds”. In:Lett. Math. Phys.107 (2017), pp. 643–658. [Wal07] Waldmann, S. Poisson-Geometrie und Deformationsquantisierung. Eine Einführung. Heidel- berg, Berlin, New York: Springer-Verlag,

work page 2017
[530]

Morita equivalence and characteristic classes of star products

[BDW12] Burzstyn, H., Dolgushev, V., and Waldmann, S. “Morita equivalence and characteristic classes of star products”. In:Crelle’s J. reine angew. Math.662 (2012), pp. 95–163. [BNW98] Bordemann, M., Neumaier, N., and Waldmann, S. “Homogeneous Fedosov Star Products on Cotangent Bundles I: Weyl and Standard Ordering with Differential Operator Representa- t...

work page arXiv 2012
[697]

Global Homotopies for Differential Hochschild Cohomologies

[Dip+24] Dippell, M., Esposito, C., Schnitzer, J., and Waldmann, S. “Global Homotopies for Differen- tial Hochschild Cohomologies”. In:Preprint arXiv:2410.15903 (Oct. 2024). [Dit99] Dito, G. “Kontsevich Star Product on the Dual of a Lie Algebra”. In: Lett. Math. Phys.48 (1999), pp. 307–322. [DK25] Dippell, M. and Kern, D. “Constraint vector bundles and re...

work page Pith review arXiv 1999
[2003]

Morita equivalence bimodules for Wick type star products

[NW03] Neumaier, N. and Waldmann, S. “Morita equivalence bimodules for Wick type star products”. In: J. Geom. Phys.47 (2003), pp. 177–196. [NW09] Neumaier,N.andWaldmann,S.“DeformationQuantizationofPoissonStructuresAssociated to Lie Algebroids”. In:SIGMA 5 (2009), p

work page 2003
[2005]

Déformations de l’Algèbre des Fonctions d’une Variété Symplectique: Compara- ison entre Fedosov et DeWilde, Lecomte

Geometry & Topology Publications, Coventry, 2011, pp. 1–107. 31 [Del95] Deligne, P. “Déformations de l’Algèbre des Fonctions d’une Variété Symplectique: Compara- ison entre Fedosov et DeWilde, Lecomte”. In:Sel. Math. New Series 1.4 (1995), pp. 667–

work page 1995
[2007]

Convergence of Star Products: From Examples to a General Framework

[Wal19] Waldmann, S. “Convergence of Star Products: From Examples to a General Framework”. In: EMS Surv. Math. Sci.6 (2019), pp. 1–31. [Wil07] Willwacher, T. “A counterexample to the quantizability of modules”. In: Lett. Math. Phys. 81.3 (2007), pp. 265–280. 33

work page 2019

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [1]

Quantum Me- chanics as a Deformation of Classical Mechanics

[Bay+77] Bayen, F., Flato, M., Frønsdal, C., Lichnerowicz, A., and Sternheimer, D. “Quantum Me- chanics as a Deformation of Classical Mechanics”. In:Lett. Math. Phys.1 (1977), pp. 521–

work page 1977

[2] [5]

Convergence of the Gutt Star Product

[ESW17] Esposito, C., Stapor, P., and Waldmann, S. “Convergence of the Gutt Star Product”. In:J. Lie Theory27 (2017), pp. 579–622. [ESW19] Esposito, C., Schmitt, P., and Waldmann, S. “Comparison and Continuity of Wick-type Star Products on Certain Coadjoint Orbits”. In:Forum Math.31.5 (2019), pp. 1203–1223. [Fed94] Fedosov, B. V. “A Simple Geometrical Con...

work page 2017

[3] [74]

Characterstic classes of star products on Marsden-Weinstein reduced symplectic manifolds

[Rei17] Reichert, Thorsten. “Characterstic classes of star products on Marsden-Weinstein reduced symplectic manifolds”. In:Lett. Math. Phys.107 (2017), pp. 643–658. [Wal07] Waldmann, S. Poisson-Geometrie und Deformationsquantisierung. Eine Einführung. Heidel- berg, Berlin, New York: Springer-Verlag,

work page 2017

[4] [530]

Morita equivalence and characteristic classes of star products

[BDW12] Burzstyn, H., Dolgushev, V., and Waldmann, S. “Morita equivalence and characteristic classes of star products”. In:Crelle’s J. reine angew. Math.662 (2012), pp. 95–163. [BNW98] Bordemann, M., Neumaier, N., and Waldmann, S. “Homogeneous Fedosov Star Products on Cotangent Bundles I: Weyl and Standard Ordering with Differential Operator Representa- t...

work page arXiv 2012

[5] [697]

Global Homotopies for Differential Hochschild Cohomologies

[Dip+24] Dippell, M., Esposito, C., Schnitzer, J., and Waldmann, S. “Global Homotopies for Differen- tial Hochschild Cohomologies”. In:Preprint arXiv:2410.15903 (Oct. 2024). [Dit99] Dito, G. “Kontsevich Star Product on the Dual of a Lie Algebra”. In: Lett. Math. Phys.48 (1999), pp. 307–322. [DK25] Dippell, M. and Kern, D. “Constraint vector bundles and re...

work page Pith review arXiv 1999

[6] [2003]

Morita equivalence bimodules for Wick type star products

[NW03] Neumaier, N. and Waldmann, S. “Morita equivalence bimodules for Wick type star products”. In: J. Geom. Phys.47 (2003), pp. 177–196. [NW09] Neumaier,N.andWaldmann,S.“DeformationQuantizationofPoissonStructuresAssociated to Lie Algebroids”. In:SIGMA 5 (2009), p

work page 2003

[7] [2005]

Déformations de l’Algèbre des Fonctions d’une Variété Symplectique: Compara- ison entre Fedosov et DeWilde, Lecomte

Geometry & Topology Publications, Coventry, 2011, pp. 1–107. 31 [Del95] Deligne, P. “Déformations de l’Algèbre des Fonctions d’une Variété Symplectique: Compara- ison entre Fedosov et DeWilde, Lecomte”. In:Sel. Math. New Series 1.4 (1995), pp. 667–

work page 1995

[8] [2007]

Convergence of Star Products: From Examples to a General Framework

[Wal19] Waldmann, S. “Convergence of Star Products: From Examples to a General Framework”. In: EMS Surv. Math. Sci.6 (2019), pp. 1–31. [Wil07] Willwacher, T. “A counterexample to the quantizability of modules”. In: Lett. Math. Phys. 81.3 (2007), pp. 265–280. 33

work page 2019