Egorov's theorem in the Weyl--H\"ormander calculus

Antoine Prouff

REVIEW 3 major objections 4 minor 12 references

Egorov's theorem in the Weyl--H\"ormander calculus

T0 review · 3 major / 4 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read This paper proves a global Egorov theorem for the Weyl–Hörmander calculus: under explicit assumptions on a Hamiltonian and an admissible phase-space metric, conjugation by the Schr\"odinger propagator preserves the pseudo-differential…

desk verdict A genuine and carefully proved Egorov theorem in the Weyl–Hörmander calculus, with one load-bearing technical assumption that the author himself flags as unnatural; worth refereeing seriously. read the letter →

arxiv 2412.04320 v1 pith:R7NNE5EI submitted 2024-12-05 math.AP

Antoine Prouff This is my paper · ORCID

classification math.AP MSC 35S3081Q2081S3035S0547D06

keywords Egorov'stheoremMicrolocalanalysisQuantum-classicalcorrespondenceWeyl–HörmandercalculusMetricsonthephasespaceEhrenfesttimePseudo-differentialoperatorsDysonseries

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

The paper proves a global version of Egorov's theorem for Schr\"odinger-type propagators on Euclidean space, working inside the metric-based Weyl–H\"ormander calculus. The claim is that for a large class of Hamiltonians and admissible phase-space metrics, conjugating any pseudo-differential observable by the unitary evolution $e^{itP}$ produces another pseudo-differential operator whose full Weyl symbol is controlled in a naturally transported symbol class $S(m(t),g(t))$, for times up to any fixed time plus half an explicitly defined Ehrenfest time. The payoff is uniformity: the seminorm estimates do not deteriorate with time, and the asymptotic expansion of the symbol is given to all orders, with corrections ordered by powers of the metric's gain function. The same theorem covers Schr\"odinger, half-wave, and transport evolutions in a single setting. The proof is driven by a stronger result on the propagation of metric-adapted partitions of unity, from which the symbol-class statement is obtained by superposition.

What carries the argument

The machinery is the family of evolved metrics $g(t)=e^{2(\Lambda+2\Upsilon)|t|}g$ together with the Dyson expansion of the quantum dynamics. The Weyl\/Moyal calculus splits the quantum generator $i(p\#a-a\#p)$ into the Hamiltonian vector field $H_p$ plus a third-order remainder $H_p^{(3)}$; the paper writes $e^{t\mathcal H_p}$ as the classical pullback $e^{tH_p}$ plus simplex integrals of $H_p^{(3)}$ conjugated by the classical flow. The key estimate is that $H_p^{(3)}$ gains two powers of the gain function $h_g$ and a negative power of the temperance weight $\theta_g$, so each Dyson term is genuinely higher order. The Ehrenfest time $T_E=\frac{1}{2(\Lambda+2\Upsilon)}\log(1/h_g)$ is exactly the time beyond which the evolved metric $g(t)$ violates the uncertainty principle $h_{g(t)}\le1$. Theorem II, the propagation of quantum partitions of unity, is the stronger result from which Theorem I is derived by superposition.

What would settle it

Take a semiclassical Schr\"odinger Hamiltonian $p=\frac12|\xi|^2+V(x)$ with a bounded-below subquadratic potential such as $V(x)=\langle x\rangle^{4/3}$, and the metric $g=dx^2+\hbar^2d\xi^2$, for which all assumptions of Theorem I hold with $\Lambda\approx\hbar$ and $\Upsilon=0$. Numerically solve the Moyal evolution equation $\partial_t a=i(p\#a-a\#p)$ for a confined initial symbol up to the predicted time $|\tau|\le\varepsilon\log(1/\hbar)$; Theorem I predicts uniform $S(1,g_\hbar)$ seminorm bounds for the whole family of symbols, so divergence of any seminorm as $\hbar\to0$ before that time would falsify the uniformity statement.

Watch

Extended reading notes

Core claim

The central claim is that, under Assumptions A and B, for any admissible metric $g$ and $g$-admissible weight $m$, the quantum transport map $e^{t\mathcal H_p}$ sends $S(m,g)$ into $S(m(t),g(t))$ for all $|t|\le T_0+\frac12 T_E$, where $g(t)=e^{2(\Lambda+2\Upsilon)|t|}g$, $m(t)=e^{tH_p}m$, and $T_E$ is the Ehrenfest time. The transported symbol has the asymptotic expansion $e^{t\mathcal H_p}a\sim e^{tH_p}a+\sum_{j\ge1}E_j(t)a$, with explicit Dyson-integral operators $E_j(t)$, and the $j$-th correction is controlled in the refined class $S(m(t)h_{g(t)}^{2j},g(t))$. This is more than a leading-order approximation: the remainder is itself a pseudo-differential operator in the expected class. The proof first establishes the stronger Theorem II: quantum evolution sends $g$-confined symbols centered at $\rho_0$ to $g(t)$-confined symbols centered at $\phi_{-t}(\rho_0)$, with confinement radius growing at a controlled exponential rate, uniformly over phase space. Theorem I follows by decomposing arbitrary symbols as superpositions of confined symbols.

Load-bearing premise

The load-bearing premise is a bound on how fast the phase-space metric can stretch along the Hamiltonian flow: for every starting point and every time, $g_{\phi_t(\rho)}$ may be at most a fixed constant times $e^{2\Upsilon|t|}g_\rho$, and the proof's control of the temperance weight relies on this; the author flags it as convenient rather than natural.

Editorial extensions

If this is right

For the flat Schr\"odinger case with subquadratic, bounded-below potentials, Theorem I gives a global Egorov theorem with an Ehrenfest time of order $\hbar^{-1}\log(1/\hbar)$ in the standard semiclassical time variable: symbols in $S(1,g_\hbar)$ remain in bounded subsets of the evolved symbol class for times $|\tau|\le\varepsilon\log(1/\hbar)$.
For half-wave evolution on a curved metric with bounded geometry, the theorem recovers and extends the standard $S^n_{\rho,\delta}$ Egorov estimates, reaching the boundary $\rho=1/2$, and allows general admissible weights rather than only powers of $\langle\xi\rangle$.
For transport equations generated by bounded vector fields preserving a density, observables depending on both position and momentum evolve pseudo-differentially on fixed time intervals; the paper also covers a family of anisotropic metrics, although the fully flow-adapted anisotropic metrics are not reproduced.
The Schr\"odinger propagator itself is shown to map the Schwartz space to itself continuously, and the quantum dynamics acts continuously on Schwartz functions and distributions on phase space.
Because the asymptotic expansion is available to all orders, the semiclassical approximation can be pushed to arbitrarily high powers of the Planck-scale gain function, which is what allows the extension to a fraction of the Ehrenfest time as $\hbar\to0$.

Reading between the lines

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

The author explicitly leaves open whether Assumption B(iii) can be replaced by geodesic temperance; if it can, the theorem would extend to metrics whose distortion under the flow is not uniformly exponential, and the Ehrenfest time would be measured by a genuinely intrinsic expansion rate.
Theorem II, rather than the symbol-class theorem alone, is the natural tool for proving observability and stabilization estimates without truncating to an energy shell, since it controls the evolution of second-microlocal symbols at all phase-space scales; the paper signals control-theoretic applications as its motivation but does not develop them.
The Dyson structure suggests a general principle: whenever the Weyl-calculus remainder gains two powers of the gain function and the flow expands the metric at a uniform exponential rate, an all-order Egorov theorem should hold by the same simplex-integral estimates, possibly including non-self-adjoint or dissipative generators.

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

Read Prouff's paper. It proves a real theorem: global Egorov conjugations for a wide class of Hamiltonians in the Weyl–Hörmander calculus, with a quantified Ehrenfest time, a full asymptotic expansion, and a stronger, genuinely new Theorem II on propagation of quantum partitions of unity. The applications to Schrödinger, half-wave, and transport evolutions are concrete and show the framework earns its keep. The proof is long but coherent: Dyson series, Beals' theorem, and the pseudo-differential calculus estimates in the appendices are all laid out with explicit assumptions. No circularity; the cited results are standard and the few self-citations are to the author's own applications, not to the theorem being proved.

The soft spots are real but not fatal. The main one is Assumption B(iii), the uniform control of the metric along the Hamiltonian flow. It is used essentially to control the temperance weight under the flow (Lemma 5.1, then Lemma 5.4, Proposition 5.5, and the confined-symbol estimates in Section 7). The author says in Remark 1.30 that this assumption is 'not so natural' and convenient rather than necessary, and points to geodesic temperance as an alternative. The stress-test note is accurate: if B(iii) fails, the gain from h_g^{3/2} may no longer dominate the nonlocal part of the Moyal commutator. This does not invalidate the theorem as stated, but it means the advertised generality is hostage to a non-intrinsic condition. The 1/2 factor in the Ehrenfest time is also acknowledged as technical. Both are places where a careful referee should push, but they do not undermine the central argument.

I did not machine-check the proof, so cautious on correctness risk, but the structure is honest and detailed. The paper is for microlocal analysts working on Egorov-type theorems, quantum-classical correspondence, or control theory applications of the Weyl–Hörmander calculus. A serious editor should send this to peer review; it deserves a careful referee even if substantial revision is likely. I would cite it if I worked in this area; I might not bring the full 90 pages to a reading group, but I would point colleagues to the introduction and Theorem II.

Referee Report

3 major / 4 minor

Summary. The paper proves a global Egorov theorem in the Weyl–Hörmander calculus on R^d. Under two sets of assumptions (A: classical and quantum well-posedness; B: compatibility of the Hamiltonian with an admissible metric g, including a Lyapunov-type bound and strong sub-quadraticity), Theorem I states that for times |t| ≤ T0 + 1/2 T_E the conjugated operator e^{itP} Op(a) e^{-itP} has Weyl symbol in S(m(t), g(t)), with an asymptotic expansion whose terms are described by a Dyson series. Theorem II is a stronger propagation result for g-uniformly confined families of symbols, and several applications are given: semiclassical Schrödinger operators, half-wave operators on curved spaces, and transport-type vector fields. The paper is largely self-contained, with detailed proofs of the pseudo-differential estimates in Appendices A–C.

Significance. If correct, the paper gives a unified and fairly general Egorov theorem that interpolates and extends several existing results: the semiclassical Egorov theorems of Bouzouina–Robert and Bambusi–Graffi–Paul, the microlocal statements of Taylor, and the Weyl–Hörmander framework used by Bony. The explicit Dyson expansion and the quantitative Ehrenfest time are useful and go beyond many previous statements. The proof strategy is coherent: the classical flow estimates are separated from the quantum commutator estimates, the main work is done in the confined-symbol theorem, and Beals' theorem is applied with careful tracking of constants. The paper also ships very detailed proofs of the auxiliary pseudo-differential calculus facts, which is a genuine strength.

major comments (3)

[§1.4, Assumption B(iii), Eq. (1.35); Remark 1.30] Assumption B(iii) is load-bearing: it is used in Lemma 5.1 to obtain the two-sided control of the temperance weight along the flow, and that estimate is fed into Proposition 5.5 and into the Beals-theorem step of Section 7.3. The author himself states in Remark 1.30 that this assumption is 'not so natural' and is convenient rather than necessary. As a referee I would ask that this issue be resolved: either prove a version under the more intrinsic geodesic temperance suggested in Remark 4.11, or present B(iii) as a genuinely central structural hypothesis and moderate the abstract's claim of 'mild assumptions'. Since all three applications in Section 1.7 have Υ = 0, the paper does not currently demonstrate that the general Υ term is natural or necessary.
[Theorem I, Eq. (1.37)–(1.38); Remark 1.26; Section 8] The statement allows T up to T0 + 1/2 T_E, but the uniform admissibility of the family g(t) is established only for |t| ≤ T_E (Remark 1.26), and Theorem II is proved only for T ≤ 1/2 T_E. The transition from Theorem II to the full range in Theorem I needs an explicit argument, presumably a rescaling of the metric as indicated in Remark 1.38. This is particularly delicate when T_E = 0, since then the range T ≤ T0 is nontrivial while g(t) may violate the uncertainty principle for t ≠ 0. Please spell out this rescaling step in Section 8, or otherwise state precisely which of the two time ranges is actually covered by the proof.
[§1.5.3, Eq. (1.37)] The theorem reaches only 1/2 T_E, while the standard Ehrenfest time is T_E. The factor 1/2 is acknowledged as technical, but it directly affects the paper's headline claim of 'quantifying an Ehrenfest time'. I recommend stating this limitation explicitly in the abstract and in the introduction, rather than only in Remark 1.5.3, and adding a remark on whether the factor 1/2 is expected to be improvable under the same assumptions.

minor comments (4)

[Theorem I, final dependence paragraph] The dependence 'on any constant c such that Λ ≥ c h_g' is unusual when Λ = 0, since no positive c exists. Please clarify whether Theorem I is asserted in the integrable case and how the constants in (1.38)–(1.42) depend on T0 and on c in that case.
[Lemma 5.1] There is a typographical corruption in the display: 'a𝑛𝑑' should read 'and'. Throughout the paper the background Euclidean metric in Definition 1.12 is denoted by the same symbol g as the admissible metric; this is a frequent source of confusion and should be renamed, for instance g0.
[Corollary 1.40, proof in §7.4] The proof applies Theorem II to the Wigner symbol u0 > u0, which is a Schwartz function and therefore belongs to every Conf^g_{r0}(ρ0). This is correct, but the uniformity in ρ0 of the confinement seminorms of u0 > u0 is not stated explicitly; a one-sentence justification would help.
[§1.5.4] The discussion of the semiclassical regime posits that the rescaled Lyapunov exponent is independent of ℏ and that c in (1.49) is uniform in ℏ, but it does not state how the constants in Theorem I behave if Λ is much smaller than h_g. Please add a remark on the non-uniform, fixed-T0 case.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the Egorov theorem is proved from explicit assumptions via pseudo-differential calculus; self-citations are non-load-bearing.

full rationale

The paper's central claim is a theorem with a proof, not a data-derived prediction. Theorem I is deduced from Theorem II on propagation of confined families, whose proof uses the Weyl–Hörmander pseudo-differential calculus (Lerner, Hörmander, Beals), the Dyson expansion, and explicit quantitative assumptions on the Hamiltonian and metric. The only appearances of the author's own prior work, [Pro23] and [Pro24], are as applications or motivation, not as premises of the proof, and no uniqueness theorem or ansatz is imported from those papers. Definition 1.2 does define e^{tH_p}a as the Weyl symbol of e^{itP}Op(a)e^{-itP}, but Theorem I is not tautological: it proves that this symbol belongs to S(m(t),g(t)) and satisfies the nontrivial asymptotic expansion e^{tH_p}a + \sum E_j(t)a. The Dyson-series identity (Proposition 1.18) is an algebraic consequence of the group law with H_p = H_p + H_p^{(3)}, and the subsequent estimates follow from pseudo-differential calculus and the stated assumptions. The flagged Assumption B(iii), Eq. (1.35), is an assumption rather than a consequence; the author explicitly calls it convenient rather than necessary in Remark 1.30 and discusses geodesic temperance as an alternative, which is a limitation of scope rather than circularity. No fitted parameter is renamed as a prediction, and no load-bearing step reduces by construction to its own input.

Assumptions & free parameters 0 free parameters · 8 assumptions · 0 invented entities

The main theorem is an a priori estimate, so there are no fitted numbers. The load-bearing axioms are Assumptions A and B, which impose regularity and boundedness conditions on the Hamiltonian and the phase-space metric. In particular, Assumption B(iii) and the lower bound on Λ are flagged by the author as technical.

assumptions (8)

domain assumption Hamiltonian flow is complete and Op(p) is essentially self-adjoint (Assumption A)
Section 1.4. Needed to define classical and quantum dynamics for all times.
domain assumption Lyapunov exponent Λ := 1/2 sup |L_{H_p} g|_g is finite (Assumption B(i))
Eq. (1.33). Controls exponential expansion of the flow.
ad hoc to paper Strong sub-quadraticity: ∇^3 p ∈ S((θ_g h_g^{1/2})^{-ε} (h_g/h_g)^{-3}, g) ∩ S((h_g/h_g)^{-1}, g) for some ε∈(0,1/2]
Assumption B(ii), Eq. (1.34). Required for the gain in the Dyson expansion; author calls it 'relatively strong'.
ad hoc to paper Metric control along flow: g_{φ_t(ρ)} ≤ C_Υ^2 e^{2Υ|t|} g_ρ for all ρ,t
Assumption B(iii), Eq. (1.35). Used to control temperance weight along flow; author notes it is 'not so natural' (Remark 1.30).
ad hoc to paper There exists c>0 such that Λ ≥ c h_g
Eq. (1.49). Used to make constants uniform in time; dependence on c is tracked and degenerates as c→0.
standard math Weyl-Hörmander pseudo-differential calculus estimates (Propositions 2.2, 2.4)
Proven in Appendix A using standard techniques; imported from the theory of pseudo-differential operators.
standard math Beals' characterization of pseudo-differential operators and Calderón-Vaillancourt theorem
Used to prove the conjugated operator is pseudo-differential and to obtain L^2 bounds.
standard math Existence of g-partition of unity with uniform seminorm estimates (Proposition 1.31)
Background result imported from Lerner [Ler10, Theorem 2.2.7] with an added estimate (1.78) derived in the paper.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Egorov's theorem in the Weyl--H\"ormander calculus." pith.science (2026). https://pith.science/paper/R7NNE5EI

@misc{pith2026241204320,
  author       = {Pith},
  title        = {Pith review of: Egorov's theorem in the Weyl--H\"ormander calculus},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/R7NNE5EI}},
  note         = {Machine review of arXiv:2412.04320}
}

read the original abstract

We prove a general version of Egorov's theorem for evolution propagators in the Euclidean space, in the Weyl--H\"ormander framework of metrics on the phase space. Mild assumptions on the Hamiltonian allow for a wide range of applications that we describe in the paper, including Schr\"odinger, wave and transport evolutions. We also quantify an Ehrenfest time and describe the full symbol of the conjugated operator. Our main result is a consequence of a stronger theorem on the propagation of quantum partitions of unity.

Figures

Figures reproduced from arXiv: 2412.04320 by the authors.

**Figure 1.** Level sets of the classical Hamiltonian 𝑝 defined in dimension 1 by (1.57), with a confining potential 𝑉 and ∇𝛽 constant. Dotted lines represent the unit boxes of the metric introduced in (1.59) 𝑔1 = γ ⊕ γ −1 (i.e. the product of unit balls of γ and γ −1 ). These boxes are squares with sidelength ≈ 1 that saturate the uncertainty principle ℎ𝑔1 ≤ 1. So here we let 𝛾 be a smooth, non-necessarily flat, Riemannian metri… view at source ↗

**Figure 2.** Level sets of the classical Hamiltonian 𝑝(𝑥, 𝜉) = ⟨𝜉⟩𝛾−1 . Dotted lines represent unit boxes of the metric 𝑔 defined in (1.63). They correspond to products of unit balls of 𝛾1 and 𝛾2 associated with the decomposition 𝑔 = 𝛾1 ⊕ 𝛾2 in (1.63). 1.7.3. Vector fields. Our last application concerns differential operators of order 1, namely vector fields. These are quite different from the previous cases for several reasons.… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

12 extracted references · 10 canonical work pages

[8]

Classical-Quantum correspondence in Lindblad evolution

[GHZ24] Jeffrey Galkowski, Zhen Huang, and Maciej Zworski. Classical-quantum correspondence in Lindblad evolution. arXiv:2403.09345 [math.AP] ,

work page Pith review arXiv
[9]

[Hör67] LarsHörmander.Pseudo-differentialoperatorsandhypoellipticequations.In SingularIntegrals(Proc.Sympos.PureMath., Chicago, Ill., 1966), pages 138–183. Amer. Math. Soc., Providence, RI,

work page 1966
[12]

Uniform stability of the damped wave equation with a confining potential in the Euclidean space

To appear inAnalysis & PDE. [Pro24] Antoine Prouff. Uniform stability of the damped wave equation with a confining potential in the Euclidean space. arXiv:2406.17358 [math.AP] ,

work page Pith review arXiv
[1966]

arXiv:2307.00839 [math.AP] ,

[Pro23] AntoineProuff.ObservabilityoftheSchrödingerequationwithsubquadraticconfiningpotentialintheEuclideanspace. arXiv:2307.00839 [math.AP] ,

arXiv
[1977]

Characterization of pseudodifferential operators and applications

[Bea79] Richard Beals. Correction to: “Characterization of pseudodifferential operators and applications” (Duke Math. J.44 (1977), no. 1, 45–57).Duke Math. J., 46(1):215,

work page 1977
[1989]

Équations aux Dérivées Partielles

[Bon94] Jean-MichelBony.OpérateursintégrauxdeFourieretcalculdeWeyl-Hörmander(casd’unemétriquesymplectique).In Journées “Équations aux Dérivées Partielles” (Saint-Jean-de-Monts, 1994), pages Exp. No. IX,

work page 1994
[1994]

Weyl quantization and Fourier integral operators

[Bon96] Jean-Michel Bony. Weyl quantization and Fourier integral operators. InPartial differential equations and mathematical physics(Copenhagen,1995;Lund,1995) ,volume21of Progr.NonlinearDifferentialEquationsAppl. ,pages45–57.Birkhäuser Boston, Boston, MA,

work page 1995
[1996]

Fourier integral operators and Weyl-Hörmander calculus

[Bon97] Jean-Michel Bony. Fourier integral operators and Weyl-Hörmander calculus. InNew trends in microlocal analysis (Tokyo, 1995), pages 3–21. Springer, Tokyo,

work page 1995

Show all 12 references

[2003]

[Bon07] Jean-MichelBony.Analysemicrolocaleetéquationsd’évolution.In Séminaire: ÉquationsauxDérivéesPartielles.2006–2007 , Sémin. Équ. Dériv. Partielles, pages Exp. No. XX,

2006
[2016]

Long time energy averages and a lower resolvent estimate for damped waves.arXiv:2309.12709 [math.AP],

[Lé23] Matthieu Léautaud. Long time energy averages and a lower resolvent estimate for damped waves.arXiv:2309.12709 [math.AP],

arXiv
[2020]

Mathematical methods of classical mechanics, volume 60 ofGraduate Texts in Mathematics

[Arn89] Vladimir Igorevich Arnol′d. Mathematical methods of classical mechanics, volume 60 ofGraduate Texts in Mathematics. Springer-Verlag,NewYork,1989.Translatedfromthe1974RussianoriginalbyK.VogtmannandA.Weinstein,Corrected reprint of the second (1989) edition. [Bea74] Richa...

1989
[2023]

arXiv:2102.11196 [math.DS] ,

[FT23b] FrédéricFaureandMasatoTsujii.Micro-localanalysisofcontactAnosovflowsandbandstructureoftheRuellespectrum. arXiv:2102.11196 [math.DS] ,

arXiv

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [8]

Classical-Quantum correspondence in Lindblad evolution

[GHZ24] Jeffrey Galkowski, Zhen Huang, and Maciej Zworski. Classical-quantum correspondence in Lindblad evolution. arXiv:2403.09345 [math.AP] ,

work page Pith review arXiv

[2] [9]

[Hör67] LarsHörmander.Pseudo-differentialoperatorsandhypoellipticequations.In SingularIntegrals(Proc.Sympos.PureMath., Chicago, Ill., 1966), pages 138–183. Amer. Math. Soc., Providence, RI,

work page 1966

[3] [12]

Uniform stability of the damped wave equation with a confining potential in the Euclidean space

To appear inAnalysis & PDE. [Pro24] Antoine Prouff. Uniform stability of the damped wave equation with a confining potential in the Euclidean space. arXiv:2406.17358 [math.AP] ,

work page Pith review arXiv

[4] [1966]

arXiv:2307.00839 [math.AP] ,

[Pro23] AntoineProuff.ObservabilityoftheSchrödingerequationwithsubquadraticconfiningpotentialintheEuclideanspace. arXiv:2307.00839 [math.AP] ,

arXiv

[5] [1977]

Characterization of pseudodifferential operators and applications

[Bea79] Richard Beals. Correction to: “Characterization of pseudodifferential operators and applications” (Duke Math. J.44 (1977), no. 1, 45–57).Duke Math. J., 46(1):215,

work page 1977

[6] [1989]

Équations aux Dérivées Partielles

[Bon94] Jean-MichelBony.OpérateursintégrauxdeFourieretcalculdeWeyl-Hörmander(casd’unemétriquesymplectique).In Journées “Équations aux Dérivées Partielles” (Saint-Jean-de-Monts, 1994), pages Exp. No. IX,

work page 1994

[7] [1994]

Weyl quantization and Fourier integral operators

[Bon96] Jean-Michel Bony. Weyl quantization and Fourier integral operators. InPartial differential equations and mathematical physics(Copenhagen,1995;Lund,1995) ,volume21of Progr.NonlinearDifferentialEquationsAppl. ,pages45–57.Birkhäuser Boston, Boston, MA,

work page 1995

[8] [1996]

Fourier integral operators and Weyl-Hörmander calculus

[Bon97] Jean-Michel Bony. Fourier integral operators and Weyl-Hörmander calculus. InNew trends in microlocal analysis (Tokyo, 1995), pages 3–21. Springer, Tokyo,

work page 1995