REVIEW 2 major objections 4 minor 19 references
Fat distributions with Reeb directions need not be complex contact
T0 review · 2 major / 4 minor · reviewed 2026-08-02 · deepseek-v4-flash
Pith's one-line read A fat (4,6)-distribution with two Reeb directions need not support a complex contact structure on any open set, even up to diffeomorphism.
desk verdict The main counterexample is solid and answers Bhowmick's question; the advertised infinite-codimension conclusion is not proved, because the κ-family collapses under a z1-shift diffeomorphism. 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 machinery is the canonical almost complex structure J assigned to a fat (4,6)-distribution on an oriented 6-manifold: J is uniquely characterized by preserving the distribution, matching the orientation, and making the Levi curvature complex-bilinear. Associated to J is the obstruction tensor S(u,v)=[u,v]+J[Ju,v] mod D, whose vanishing on all pairs is equivalent to the distribution germ being complex contact. The paper constructs J explicitly from a complex root of the Levi quadratic form of the model distribution, computes it on the quotient by the Reeb directions and on the distribution itself, and then evaluates S on the pair R=∂z2, ν=∂x2+y2∂z1+y1∂z2. The z1-component of the resulting
What would settle it
Recompute the obstruction tensor directly for the distribution in Theorem 3.1, taking R=∂z2 and ν=∂x2+y2∂z1+y1∂z2, and inspect the ∂z1 coefficient A1(x2) of (2[JR,ν]−J[JR,ν]−[JR,Jν]) mod D. The paper's formula gives A1(x2)=x2·f(x2) with f zero only when x2^2=−1+√2; if a recomputation yields A1≡0 on an open interval inside (−1,1), the main theorem is false.
Extended reading notes
Core claim
The paper's central object is the rank-4 distribution (R^6,D) cut out by the 1-forms λ1 = dz1 − y1 dx1 − y2 dx2 − (x2^3/3 + x2 + 2x1) dy1 and λ2 = dz2 − y2 dx1 − y1 dx2. This distribution is fat on the open slab |x2|<1 and admits two commuting Reeb directions ∂z1, ∂z2. For any fat distribution on an oriented 6-manifold there is a canonical almost complex structure; the paper computes it for this example and evaluates the obstruction tensor S, whose vanishing is exactly the condition that the germ be induced by a complex contact structure. The relevant component of S(R,ν) is shown to be a nonzero function of x2 on every open interval, so no germ can be complex contact. A diffeomorphism from t
Load-bearing premise
The proof rests on the correctness of the explicit formula for the canonical almost complex structure and on the resulting coefficient A1(x2) being nonzero on every interval; if those bracket computations contain an algebraic slip, the claim that S never vanishes is not established.
Editorial extensions
If this is right
- Fat (4,6)-distributions with Reeb directions strictly contain the class that comes from complex contact structures, already at the level of germs.
- Germs of horizontal immersions into such distributions need not be locally equivalent to 1-jet prolongations of holomorphic maps, since the complex-contact model is no longer forced.
- At every point of every 6-manifold, the space of complex-contact distribution germs has infinite codimension inside the space of fat distribution germs with Reeb directions.
- The canonical almost complex structure and the tensor S give an effective local certificate: to check whether a fat germ is complex contact, one computes whether S vanishes.
Reading between the lines
- The explicit bracket calculation suggests the non-vanishing of S is stable: because A1(x2) only vanishes at isolated points, a small C∞ perturbation of the 1-forms that keeps the distribution fat will generically preserve the obstruction, matching the paper's infinite-codimension conclusion.
- One could test the same method on distributions obtained by replacing the polynomial 3−2x2^2−x2^4 in the definition of t with nearby positive polynomials; if a similar coefficient function remains nonzero on every interval, the counterexample extends to a whole neighbourhood in the space of fat distributions.
- The hyperbolic analogue is left open: whether a hyperbolic (4,6)-distribution with Reeb directions must be a product of two real contact germs. The same obstruction-tensor strategy, if adapted, could settle it.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper studies the converse of the known implication ``complex contact structure on a real 6-manifold gives a fat (4,6)-distribution with two Reeb directions.'' It constructs an explicit pair of 1-forms on R^6 whose kernel distribution is fat in |x_2|<1, admits two Reeb directions, and whose Čap--Eastwood tensor S is shown to be nonvanishing on every open set. By Theorem 2.13 this means no germ is induced by a complex contact structure. The construction is then globalized by a diffeomorphism to all of R^6 (Theorem 3.13). Section 4 claims an infinite-dimensional family of such germs and concludes infinite codimension of complex-contact germs among fat germs with Reeb directions (Theorems 4.1 and 1.3). The main counterexample is explicit and appears sound; the infinite-codimension part does not, as explained below.
Significance. If the main counterexample stands, it settles Bhowmick's Question 1.1 in the negative and shows that fat distributions with Reeb directions form a strictly larger class than those underlying complex contact structures, even at the germ level. The paper has real strengths: the construction is fully explicit; the fatness, Reeb, and nonvanishing-S computations are written out; and the globalization via tan((pi/2)x_2) is transparent. I verified that the alleged algebraic error in Lemma 3.10 is not an error: the coefficient -2/Δ is correct, and the sign algebra leading to Eq. (20) is consistent. The serious defect is confined to Section 4: the claimed infinite-dimensional family is actually a single diffeomorphism orbit, so Theorems 4.1 and 1.3 are not established as stated.
major comments (2)
- [Section 4, Theorem 4.1 and Theorem 1.3] The perturbation term is exact: α1_κ = α1_0 - κ(y1)dy1 = α1_0 - d(∫_0^{y1} κ(s)ds). Therefore the explicit global diffeomorphism Φ(x1,x2,y1,y2,z1,z2) = (x1,x2,y1,y2,z1 - ∫_0^{y1} κ(s)ds, z2) satisfies Φ^*λ1 = α1_κ and Φ^*λ2 = α2. Hence every (R^6,H_κ) in the family is diffeomorphic to the original (R^6,D). Since non-existence of a complex contact structure is diffeomorphism-invariant, condition (iii) gives no new information. More importantly, a family contained in one Diff-orbit cannot establish that complex-contact germs have infinite codimension in the space of fat germs with Reeb directions. The proof's assertion that ``the differentials have identical expressions and therefore the same proof applies'' misses this exactness issue. This is load-bearing for Theorem 1.3, which is currently unsupported.
- [Section 4 proof structure] Even if the family were nontrivial, the manuscript does not define the relevant topology or notion of codimension on the space of germs, nor does it explain why a family parametrized by an infinite-dimensional function space implies infinite codimension rather than merely a large family of equivalence classes. The exactness defect makes this point moot for the present family, but the paper should either supply a rigorous definition and argument or state the weaker result that there exist infinitely many (non-equivalent) examples.
minor comments (4)
- [Proposition 3.12] The vector denoted η in the first paragraph and in Eq. (18)-(19) is the ν defined just before: ν = ∂_{x2} + y2∂_{z1} + y1∂_{z2}. Please use one symbol throughout.
- [Lemma 3.11] The proof of Lemma 3.11 is phrased slightly informally: ``A1 = x2 f(x2), and x2 only vanishes at the origin whereas f vanishes at most in a finite set, so A1 cannot vanish on an open set.'' Since a product of two nonzero functions can vanish at isolated points, the argument is valid, but it would be clearer to state explicitly that an open interval would have to be contained in the zero set of at least one factor.
- [Theorem 3.13] In the proof, after defining β1, β2, the displayed formulas for (23)-(24) are labelled as λ1, λ2; this is correct but the preceding sentence could confuse the reader. Also, the diffeomorphism ϕ maps {|x2|<1} to R^6, so the phrase ``global distribution'' is justified, but this should be stated explicitly as a pullback of the fat region.
- [Theorem 1.3] The statement says ``for every point p in M'' but the proof only transfers the origin germ by a local chart. This is fine, but the theorem assumes M is a smooth 6-manifold; no orientation issue is discussed. Since the Čap--Eastwood construction in the paper uses an oriented manifold, the hypotheses should be stated precisely (e.g., oriented M).
Circularity Check
No significant circularity: the central counterexample is computed from an external Čap–Eastwood obstruction tensor and explicit parameter-free formulae; self-citations are contextual only.
full rationale
The derivation chain is self-contained. Theorem 1.2/3.13 is proved by direct computation of the Čap–Eastwood tensor S (Eq. 2) for an explicit distribution, with the identifying criterion (S vanishes iff the germ is complex-contact) imported from the external theorem of Čap and Eastwood [6] (Theorem 2.12/2.13). Lemmas 3.6, 3.8, 3.10, and Proposition 3.12 are explicit algebraic and analytic computations with no fitted parameters and no appeal to the target conclusion; the globalisation in Theorem 3.13 is a direct pullback by an explicit diffeomorphism. The paper’s self-citations (e.g., [13], [14], [15]) occur in background remarks on h-principles, maximal growth, and classification, and are not load-bearing for the counterexample. Even if Theorem 4.1’s infinite-codimension claim has a separate correctness concern—the κ-perturbations may all be pullbacks of D by a z1-shift—that is a mathematical gap, not a circularity: it does not make the main construction depend on its own inputs. Accordingly, no circular step is exhibited.
Assumptions & free parameters
free parameters (1)
- perturbation function κ(y1) =
arbitrary κ ∈ C∞(R,R)
assumptions (3)
- domain assumption Čap–Eastwood criterion: on an oriented fat (4,6)-distribution, the canonical J exists and S=0 iff the germ is induced by a complex contact structure.
- standard math Fatness/ellipticity is checked by positivity of det(M_qD)=4-(x2^2+1)^2 on |x2|<1.
- domain assumption Fatness, Reeb directions, and complex-contactability are preserved by diffeomorphisms.
Cite this review
Pith. "Pith review of Fat distributions with Reeb directions need not be complex contact." pith.science (2026). https://pith.science/paper/IKK25PMW
@misc{pith2026260318808,
author = {Pith},
title = {Pith review of: Fat distributions with Reeb directions need not be complex contact},
year = {2026},
howpublished = {\url{https://pith.science/paper/IKK25PMW}},
note = {Machine review of arXiv:2603.18808}
}
abstract
It is well known that every complex contact $3$-manifold, when regarded as a real manifold, gives rise to a fat $(4,6)$-distribution that admits two Reeb directions. Nonetheless, it was an open question whether the converse was true. This was not known even at the level of germs. The present work completely answers this question in the negative. We construct the first example of a fat distribution with two Reeb directions that does not support a complex contact structure anywhere, not even locally nor up to diffeomorphism. This result answers an open question by Aritra Bhowmick.
Figures
Reference graph
Works this paper leans on
-
[1]
Alarc´ on, F
A. Alarc´ on, F. Forstneriˇ c, F. J. L´ opez.Holomorphic Legendrian curves. Compositio Mathematica. 2017;153(9):1945-1986
2017
-
[2]
Bande, A
G. Bande, A. Hadjar.Contact pairs. Tohoku Math. J. 57 (2005), 247–260
2005
-
[3]
Bhowmick.On horizontal immersions of discs in fat distributions of type(4,6).Journal of Topology and Analysis 2024 16:01, 125-153
A. Bhowmick.On horizontal immersions of discs in fat distributions of type(4,6).Journal of Topology and Analysis 2024 16:01, 125-153
2024
-
[4]
Bhowmick.Horizontal, Contact and Partially Horizontal Immersions in Fat Distributions
A. Bhowmick.Horizontal, Contact and Partially Horizontal Immersions in Fat Distributions. Doctoral Theses. Indian Statistical Institute. 569. 2022. F AT DISTRIBUTIONS WITH REEB DIRECTIONS NEED NOT BE COMPLEX CONTACT 23
2022
-
[5]
Bhowmick, M
A. Bhowmick, M. Datta.Existence of horizontal immersions in fat distributions. International Journal of Mathematics 2023 34:10
2023
-
[6]
Geometry and Physics
A. ˇCap, M. Eastwood.Some special geometry in dimension six. Proceedings of the 22nd Winter School “Geometry and Physics” pp. 93-98. Palermo: Circolo Matematico di Palermo, 2003
2003
-
[7]
Cieliebak, Y
K. Cieliebak, Y. Eliashberg, N. Mishachev.Introduction to the h-principle. 2nd Edition. Graduate Studies in Mathematics, 239. American Mathematical Society,
-
[8]
Console, A
S. Console, A. Fino, E. Samiou. The moduli space of six-dimensional two-step nilpotent Lie algebras. Annals of Global Analysis and Geometry, 27(1), 17-32. 2005
2005
Show all 19 references
-
[9]
Fern´ andez,´A
E. Fern´ andez,´A. del Pino, W. ZhouThe h-principle fails for prelegendrians in corank 2 fat distributions.Preprint. arXiv:2511.17780
-
[10]
Ge.Betti numbers, characteristic classes and sub-Riemannian geometry.Illinois Journal of Mathematics 36 (3), 372-403, 1992
Z. Ge.Betti numbers, characteristic classes and sub-Riemannian geometry.Illinois Journal of Mathematics 36 (3), 372-403, 1992
1992
-
[11]
Geiges.An Introduction to Contact Topology.Cambr
H. Geiges.An Introduction to Contact Topology.Cambr. Studies in Adv. Math. 109. Cambr. Univ. Press 2008
2008
-
[12]
Gr` acia, ´A
X. Gr` acia, ´A. Mart ´ ınez-Mu˜ noz, X. Rivas.Pairs of differential forms: a framework for precontact geometry. Preprint: arXiv:2602.04882
-
[13]
Mart ´ ınez-Aguinaga.Existence and classification of maximal growth distributions
J. Mart ´ ınez-Aguinaga.Existence and classification of maximal growth distributions. Rev. Mat. Iberoam. 42 (2026), no. 1, pp. 31–74
2026
-
[14]
Mart ´ ınez-Aguinaga,´A
J. Mart ´ ınez-Aguinaga,´A. del Pino.Convex integration with avoidance and hyperbolic (4,6) distributions. Preprint. arXiv:2112.14632
-
[15]
Mart ´ ınez-Aguinaga,´A
J. Mart ´ ınez-Aguinaga,´A. del Pino.Classification of tangent and transverse knots in bracket- generating distributions. Preprint. arXiv:2210.00582
-
[16]
Montgomery.Generic Distributions and Lie Algebras of Vector Fields
R. Montgomery.Generic Distributions and Lie Algebras of Vector Fields. Journal of Differential Equations, pp. 387-393. Vol. 103, Issue 2, 1993
1993
-
[17]
del Pino, T
´A. del Pino, T. Shin.Microflexiblity and local integrability of horizontal curves. Math. Nachr. 297 (2024), 3252–3287
2024
-
[18]
del Pino, L
´A. del Pino, L. Toussaint.Wrinkling h-principles for integral submanifolds of jet spaces. Preprint arXiv:2112.14720
-
[19]
De Zanet.Generic one-step bracket-generating distributions of rank four
C. De Zanet.Generic one-step bracket-generating distributions of rank four. Archivum Mathematicum 051.5 (2015): 257-264. Universidad Complutense de Madrid, Departamento de ´Algebra, Geometr ´ıa y Topolog´ıa, F acultad de Ciencias Matem´aticas. 28040 Madrid, Spain Email address...
2015
Reviewed August 2, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.