Pith. sign in

REVIEW 4 major objections 4 minor 23 references

Knots and non-orientable surfaces in 3-manifolds

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

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

Pith's one-line read The paper proves that every link in a splittable 3-manifold is isotopic to a non-orientable plat closure of a surface braid, and gives explicit splittings for lens spaces L(2k,q) and trivial circle bundles Σ×S^1.

desk verdict Useful extension of plat closures to one-sided splittings, but the main proof is a sketch and Lemma 3.1 is plainly wrong; referee it, but expect major revision. read the letter →

arxiv 2502.06984 v2 pith:JUZAHABR submitted 2025-02-10 math.GT

Alessia Cattabriga , Paolo Cavicchioli , Rama Mishra , Visakh Narayanan This is my paper · ORCID
classification math.GT MSC 57K1020F3657K30
keywords non-orientableplatclosureone-sidedHeegaardsplittingsplittable3-manifoldssurfacebraidgroupslensspacestrivialcirclebundlessurfacesknotsandlinks
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 introduces a normal form for knots and links inside a broad class of closed orientable 3-manifolds: the 'splittable' ones, which decompose as a handlebody glued to the mapping cylinder of the orientating double cover of a non-orientable surface. The paper's central claim is that every link in such a manifold is isotopic to a non-orientable plat closure of a surface braid—a braid in the product layer Σ_{g-1}×I whose strand ends are capped by unlinked arcs in the handlebody and in the twisted neighborhood of the non-orientable surface. This unifies link descriptions in lens spaces L(2k,q) and in trivial circle bundles Σ_g×$S^{1}$, and it opens the door to studying links in these manifolds through surface braid groups.

What carries the argument

The central object is the non-orientable plat closure, built from a one-sided Heegaard splitting M = H_{g-1} ∪ Σ_{g-1}×I ∪ N(U_g). A braid in the surface braid group of Σ_{g-1} is closed up by two kinds of 'capping curves': unlinked boundary-parallel arcs in the handlebody H_{g-1}, and similar residual capping curves in the twisted I-bundle N(U_g) over the non-orientable surface U_g. These capping curves are the mechanism that converts a surface braid into a closed link, and the isotopy argument in Theorem 3.1 is the claim that any link can be rearranged into exactly this form by sliding arcs and critical points across the three layers.

What would settle it

Compute the relative homology H_1(N(U_g),∂N(U_g)) by a cellular or Mayer-Vietoris calculation for the Klein-bottle case U_2: if the group is Z/2 rather than Z/2⊕Z/2, then the capping-curve collection in Definition 3.1 is larger than the topology permits, and the capping step of Theorem 3.1 would need to be revised.

Watch

Extended reading notes

Core claim

The central discovery is Theorem 3.1: every link in a splittable 3-manifold M = H_{g-1} ∪_φ C(U_g) is isotopic to the non-orientable plat closure of a braid in the surface braid group of Σ_{g-1}=∂H_{g-1}. The proof decomposes M as H_{g-1} ∪ Σ_{g-1}×I ∪ N(U_g) and then slides an arbitrary link into a standard position: monotone braided strands in the middle product layer, with all critical points pushed into capping curves in the two outer pieces. The paper also shows that the Bredon-Wood embeddings in lens spaces L(2k,q) and the standard embeddings in Σ_g×$S^{1}$ induce one-sided Heegaard splittings, so the theorem applies there and explicit non-orientable plat descriptions exist for those manifolds.

Load-bearing premise

The proof's load-bearing premise is that any arc lying inside the thickened non-orientable surface can be slid, with its endpoints kept on the boundary, into a standard collection of mutually unlinked capping curves, and that the analogous slides of maxima and minima in the product layer never create obstructions; this is illustrated in Figures 8 to 10 but not proven, and the relative-homology count in Lemma 3.1 that fixes how many capping curves are needed appears to misreport the group as (Z/2)^g when its own generators would collapse to a single Z/2.

Editorial extensions

If this is right

  • Every link in every splittable 3-manifold has a non-orientable plat presentation, so surface braid groups become a common algebraic language for links in these manifolds.
  • The lens spaces L(2k,q) and the trivial circle bundles Σ_g×S^1 are all covered, and the paper gives explicit splitting surfaces and handlebody complements for the families L(2k,1) and L(4a+4,2a+1).
  • Since any closed orientable 3-manifold that contains an embedded non-orientable surface is splittable, the representation applies well beyond the worked examples.
  • The non-orientable plat closure provides a direct counterpart to the orientable plat closures associated with Heegaard splittings, setting up a framework in which link invariants could be developed from the braid algebra.

Reading between the lines

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

  • If the isotopy theorem is correct and the capping-curve count can be made rigorous, a natural next step is a Markov-type equivalence for non-orientable plat closures, which would let one compare different braid words representing the same link; the paper explicitly leaves the algebraic study to a sequel.
  • One could use this normal form to define numerical invariants, such as a minimal braid index or a plat bridge number for links in splittable manifolds, generalizing classical invariants from S^3; the paper does not explore these.
  • The validity of the construction depends on the relative-homology count in Lemma 3.1: if H_1(N(U_g),∂N(U_g)) is only Z/2 rather than the claimed (Z/2)^g, then the standard collection of capping curves is smaller than stated, and the capping step in the plat definition needs a revised description.
  • A direct computational check of the theorem on explicit knots in L(2k,1) or in Σ×S^1, converting a given diagram into the claimed plat form, would test the sliding procedure and could reveal where a rigorous lemma is missing.
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.

Referee Report

4 major / 4 minor

Summary. The paper defines a non-orientable plat closure for links in closed orientable 3-manifolds that admit a one-sided Heegaard splitting M = H_{g-1} ∪ (Σ_{g-1}×I) ∪ N(U_g), with U_g a closed non-orientable surface of genus g. Theorem 3.1 asserts that every link in such a splittable manifold is isotopic to the non-orientable plat closure of a surface braid in Σ_{g-1}×I. Sections 4 and 5 construct explicit splittings for lens spaces L(2k,1), L(4a+4,2a+1), and trivial circle bundles Σ_g×S^1 via Bredon–Wood and End embeddings. The proof of Theorem 3.1 is a figure-based isotopy sketch; Lemma 3.1, which is used to describe capping curves, is false as stated.

Significance. This is a potentially useful extension of the RP^3 plat construction of Mishra–Narayanan to the class of splittable manifolds introduced by Rubinstein. A rigorous version of Theorem 3.1 would give a uniform braid-plat normal form for links in a large family of orientable 3-manifolds and would connect with surface braid groups and skein modules. The explicit examples in Sections 4 and 5 are concrete and checkable, and the paper is honest in describing its scope. The central claim is not yet supported by a complete proof, so the paper cannot be accepted in its present form.

major comments (4)
  1. [§3, Lemma 3.1] The asserted isomorphism H_1(N(U_g), ∂N(U_g)) ≅ (Z/2)^g is false for g>1. Since N(U_g) deformation retracts to U_g, H_1(N(U_g)) ≅ Z^{g-1} ⊕ Z/2. Lefschetz duality for the orientable 3-manifold N(U_g) gives H_1(N(U_g),∂N(U_g)) ≅ H^2(N(U_g)) ≅ Ext(H_1(N(U_g)),Z) = Z/2. The proof in the paragraph around Figure 6 double-counts the unique nonzero relative class. This matters because Definition 3.1 and the main construction rely on the stated description of capping curves in N(U_g).
  2. [§3, proof of Theorem 3.1, paragraphs after Figures 8–10] The main isotopy assertion is not proved. The text states that every boundary-parallel arc in N(U_g) can be slid so that its intersection with N(U_g) is a collection of mutually unlinked capping curves and its intersection with Σ_{g-1}×I is boundary-parallel, but no argument or reference is given. This is not a standard general-position fact: a boundary-parallel arc is trivial in H_1(N(U_g),∂N(U_g)), whereas a capping curve is described as a generator, so the claimed sliding must move endpoints through Σ_{g-1}×I in a controlled way. The analogous maxima/minima slides in Figures 9 and 10 are also only schematic. Since this step is exactly what converts an arbitrary link into plat form, the proof of Theorem 3.1 is incomplete.
  3. [§3, proof of Theorem 3.1, first paragraph] The proof assumes that an arbitrary link can be considered as a union of boundary-parallel arcs in H and N(U_g) and of curves in Σ_{g-1}×I. General position gives only neatly embedded arcs; a separate argument is needed to show that every link component can be isotoped to this form before the sliding steps. This missing hypothesis is load-bearing because the subsequent figures all start from boundary-parallel arcs.
  4. [§3, Definition 3.1 and surrounding notation] The definition of non-orientable plat closure is too loose to make Theorem 3.1 precise. The braid group B_{g-1,n} is not defined, the number of ends of β and the number and placement of capping curves are not specified, and 'mutually unlinked' capping curves are explained only informally by saying that two capping curves are fibres in some parametrization of N(U_g). A rigorous statement of Theorem 3.1 needs a precise description of how the 2n ends of the braid are joined to a specified collection of capping curves in H_{g-1} and N(U_g).
minor comments (4)
  1. [Throughout] There are several typographical errors, e.g. 'One might what kinds' (Section 2), 'Asconsequenceallclosed' (after Theorem 2.2), and 'the resulting manifold is an handlebody' (Section 4.1).
  2. [References] The reference [DL15] is listed twice in the bibliography and one of the two entries should be removed.
  3. [Abstract and Theorem 2.1] The notation Z/2Z is typeset inconsistently, and the passage from Rubinstein's open-handlebody conclusion to the compact-handlebody definition of splittability should be made explicit.
  4. [Figures 8–10] Since these figures carry the main proof, they need more precise captions and a formal statement of the sliding lemma they illustrate; in their current form they are too schematic to verify the isotopy.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular reduction: Theorem 3.1 is an independent geometric construction; self-citations are definitional and non-load-bearing, while the main weaknesses are an unproved sliding assertion and an incorrect homology lemma, which are correctness issues rather than circularity.

full rationale

The central claim (Theorem 3.1) is not derived by fitting parameters or by citing a result that already contains it. The non-orientable plat closure is defined directly via capping curves and braids in Section 3, and the theorem asserts a genuine isotopy statement. The proof attempts to justify the reduction by sliding arcs (Figures 8-10); whether those slides are valid is a completeness/correctness question, not a circularity. The only self-citation, [MN23], is used to motivate the notion of residual/capping curves, but Lemma 3.1 and Figure 7 give the construction in the present paper, and Theorem 3.1 does not quote [MN23] as its justification. Rubinstein's splitting theorem [Rub78] and Bredon-Wood/End embeddability results are independent external inputs. No equation or parameter in the paper reduces to its own output; in particular, Lemma 3.1's asserted (Z/2Z)^g relative homology and the unproved 'every boundary-parallel arc can be slided to capping curves' claim are substantive mathematical gaps, but they are not cases of the conclusion being assumed by definition.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The paper relies on established theorems about one-sided Heegaard splittings and non-orientable embeddings rather than introducing new physical postulates. No free parameters are fitted. The main flaw is internal: Lemma 3.1 miscomputes a relative homology group.

assumptions (3)
  • domain assumption Rubinstein's one-sided Heegaard splitting theorem (Theorem 2.1 in the paper)
    Assumed to guarantee that every nontrivial Z/2Z homology class is represented by a non-orientable surface whose complement is an open handlebody, so that splittable manifolds exist in abundance.
  • domain assumption End's result on non-zero Z/2Z homology of embedded non-orientable surfaces (Theorem 2.2)
    Used to conclude that any closed orientable 3-manifold containing a non-orientable surface admits a one-sided splitting.
  • domain assumption Bredon-Wood classification of non-orientable surfaces in lens spaces (Theorem 4.1)
    Used in Section 4 to identify the lens spaces L(2k,q) that admit non-orientable embeddings and their minimal genus.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Knots and non-orientable surfaces in 3-manifolds." pith.science (2026). https://pith.science/paper/JUZAHABR

@misc{pith2026250206984,
  author       = {Pith},
  title        = {Pith review of: Knots and non-orientable surfaces in 3-manifolds},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JUZAHABR}},
  note         = {Machine review of arXiv:2502.06984}
}
abstract

In this article, we propose a new approach for describing and understanding knots and links in a 3-manifold through the use of an embedded non-orientable surface. Specifically, we define a plat-like representation based on this non-orientable surface. The method applies to manifolds of the form $M=\mathcal H\cup_{\varphi} \mathcal C(U)$ where $\mathcal H$ is a handlebody, $\mathcal C(U)$ is the mapping cylinder of the orientating two sheeted covering of a non-orientable closed surface $U$ and $\varphi:\partial \mathcal H\to \partial \mathcal C(U)$ is an attaching homeomorphism. We show that, by fixing such a splitting any link in the manifold can be represented as a plat-like closure of an element of the surface braid group of $\partial \mathcal H$. Manifolds of this type were extensively studied by J.H. Rubinstein \cite{rubinstein1978one}, where it is shown that any 3-manifold $M$, with a non-vanishing $H_2(M,\frac{\mathbb{Z}}{2\mathbb{Z}})$ will admit such a splitting. Thus the method is quite general. We provide explicit examples of such embeddings in lens spaces $L(2k,q)$ and the trivial circle bundles over orientable closed surfaces, $\Sigma\times S^1$

Figures

Figures reproduced from arXiv: 2502.06984 by the authors.

Figure 1
Figure 1. Mapping cylinder of the Klein bottle. Since the 3-manifolds we are interested in are orientable, the normal bundle of i(Ug) is non￾trivial. As consequence all closed, connected, orientable 3-manifolds containing an embedded closed connected non-orientable surface are splittable. Important examples of splittable 3-manifolds are lens spaces L(2k, q) and trivial circle bundles over closed orientable surfaces, Σg × S 1 … view at source ↗
Figure 2
Figure 2. Capping curves inside a handlebody. Let’s pass to the tubular neighborhood N (U). As capping curves we will consider curves that generalize the notion of residual curves introduced in [MN23] for the case of the projective space RP3 . Before defining them precisely we need a preparatory lemma. Lemma 3.1. Let Ug be a genus g closed connected non-orientable surface embedded in a closed connected orientable manifold M. … view at source ↗
Figure 3
Figure 3. Two spheres with connected with a tube [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (14 more)
Figure 4
Figure 4. Figure 4: Thickening two spheres connected by a tube. [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Thickenings of the cases g = 2 and g = 3. 6 [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: Cancellation of the cores of the cylinders. [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: A capping curve in C. In the relative homology group H1(N (Ug), ∂N (Ug)), the cores of all of the cylinders are trivial cycles, since they are homologous to a circle in ∂N (Ug) as shown in [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]
Figure 8
Figure 8. Figure 8: Sliding an arbitrary curve to obtain capping curves in [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 9
Figure 9. Figure 9: Sliding minima points in Σg−1 × I into Hg−1. We may slide any maxima point to N (Ug) so that we get two monotonic arcs in Σg−1 and a boundary parallel arc in N (Ug). Now as described above, we may again slide that boundary parallel arc in N (Ug) so that we get a collec…
Figure 10
Figure 10. Figure 10: A maxima point in Σg−1 × I. Sliding it to obtain capping curves in N (Ug) and a minima point in Σg−1 × I, then sliding the minima point from Σg−1 × I to obtain braided curves in Σg−1 × I and a capping curve in Hg−1. The following result ([BW69, Section 6]) describes t…
Figure 11
Figure 11. Figure 11: A minimal genus τ function for the case of L(2k, 1), for k odd. When k is even the label k − 2 is replaced with k − 1 and so on [PITH_FULL_IMAGE:figures/full_fig_p011_11.png]
Figure 12
Figure 12. Figure 12: The single curve of a minimal genus τ function for the case of L(2k, 1), for k odd. When k is even the label k − 2 is replaced with k − 1 and so on. By [BW69] the minimal g, obtained by varying τ , is also the minimal g for which Ug embeds into L(2k, q) so it is N(2k,…
Figure 13
Figure 13. Figure 13: The surface Fk (in blue), together with the core of the handles of V1 \N (Fk) (in green), in the case k = 4. Identifying the boundary disks of the solid cylinder D2 × I shown, for the case k = 4, in [PITH_FULL_IMAGE:figures/full_fig_p012_13.png]
Figure 14
Figure 14. Figure 14: A minimal genus τ function for the case of L(4a + 4, 2a + 1). so that b0 = a0 = 2, b1 = 0 and b2 = a2 = 2. This means that N(L(4a + 4, 2a + 1)) = 4, and the genus of the minimal genus non-orientable surface embedded in L(4a + 4, 2a + 1) is two and so it is a Klein bot…
Figure 15
Figure 15. Figure 15: The collection of curves for a minimal genus [PITH_FULL_IMAGE:figures/full_fig_p014_15.png]
Figure 16
Figure 16. Figure 16: A part of the non-orientable surface F2. 14 [PITH_FULL_IMAGE:figures/full_fig_p014_16.png]
Figure 17
Figure 17. Figure 17: The τ curves in a disk D and its complement in Σg. Choose an arbitrary section, σ : Σg → Σg×S 1 and denote with p : Σg×S 1 → Σg the projection. Let c be a circle in Σg bounding a disk D. In [PITH_FULL_IMAGE:figures/full_fig_p015_17.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

23 extracted references · 23 canonical work pages

  1. [1]

    Alexander

    James W. Alexander. A lemma on systems of knotted curves. Proceedings of the National Academy of Sciences , 9(3):93--95, 1923

    work page 1923

  2. [2]

    Hilden braid groups

    Paolo Bellingeri and Alessia Cattabriga. Hilden braid groups. Journal of Knot Theory and Its Ramifications , 21(03):1250029, 2012

    work page 2012

  3. [3]

    Joan S. Birman. On the stable equivalence of plat representations of knots and links. Canadian Journal of Mathematics , 28(2):264--290, 1976

    work page 1976

  4. [4]

    Bredon and John W

    Glen E. Bredon and John W. Wood. Non-orientable surfaces in orientable 3-manifolds. Inventiones mathematicae , 7(2):83--110, 1969

    work page 1969

  5. [5]

    The A lexander polynomial of (1,1)-knots

    Alessia Cattabriga. The A lexander polynomial of (1,1)-knots. Journal of Knot Theory and Its ramifications , 15:1119--1129, 2006

    work page 2006

  6. [6]

    An algorithmic method to compute plat-like Markov moves for genus two 3-manifolds

    Paolo Cavicchioli. An algorithmic method to compute plat-like markov moves for genus two 3-manifolds. arXiv preprint arXiv:2112.13718 , 2021

    work page Pith review arXiv 2021

  7. [7]

    A M arkov theorem for generalized plat decomposition

    Alessia Cattabriga and Bo s tjan Gabrov s ek. A M arkov theorem for generalized plat decomposition. Annali della Scuola Normale Superiore di PIsa, Classe di Scienze , 20:1273--1289, 2020

    work page 2020

  8. [8]

    The mixed Hilden braid group and the plat equivalence in handlebodies

    Paolo Cavicchioli and Sofia Lambropoulou. The mixed H ilden braid group and the plat equivalence in handlebodies. arXiv preprint arXiv:2312.10781 , 2023

    work page Pith review arXiv 2023

Show all 23 references
  1. [9]

    Extending homeomorphisms from punctured surfaces to handlebodies

    Alessia Cattabriga and Michele Mulazzani. Extending homeomorphisms from punctured surfaces to handlebodies. Topology and Its Applications , 155:610--621, 2008

    2008

  2. [10]

    Braid equivalence in 3-manifolds with rational surgery description

    Ioannis Diamantis and Sofia Lambropoulou. Braid equivalence in 3-manifolds with rational surgery description. Topology and Its Applications , 194:269--295, 2015

    2015

  3. [11]

    Topological steps toward the homflypt skein module of the lens spaces l(p, 1) via braids

    Ioannis Diamantis, Sofia Lambropoulou, and Jozef Przytycki. Topological steps toward the homflypt skein module of the lens spaces l(p, 1) via braids. Inventiones mathematicae , 25:1650084, 2016

    2016

  4. [12]

    A generalization of bridge number to links in arbitrary three-manifolds

    Helmut Doll. A generalization of bridge number to links in arbitrary three-manifolds. Mathematische Annalen , 1993

    1993

  5. [13]

    Non-orientable surfaces in 3-manifolds

    Werner End. Non-orientable surfaces in 3-manifolds. Archiv der Mathematik , 59:173--185, 1992

    1992

  6. [14]

    The homflypt skein module of the lens spaces l_ p,1

    Bo s tjan Gabrov s ek and Maciej Mroczkowski. The homflypt skein module of the lens spaces l_ p,1 . Journal of Knot Theory and Its Ramifications , 20:159--170, 2011

    2011

  7. [15]

    Generators for two groups related to the braid group

    Hugh Hilden. Generators for two groups related to the braid group. Pacific Journal of Mathematics , 59(2):475--486, 1975

    1975

  8. [16]

    Sofia Lambropoulou and Colin P. Rourke. Markov's theorem in 3-manifolds. Topology and its Applications , 78(1-2):95--122, 1997

    1997

  9. [17]

    Sofia Lambropoulou and Colin P. Rourke. Algebraic M arkov equivalence for links in three-manifolds. Compositio Mathematica , 142(4):1039--1062, 2006

    2006

  10. [18]

    U ber die freie \

    Andrei A. Markov. \"U ber die freie \"a quivalenz geschlossener z \"o pfe. Recueil Mathematique Moscou , 1:73--78, 1935

    1935

  11. [19]

    Plat closures of spherical braids in R p^3

    Rama Mishra and Visakh Narayanan. Plat closures of spherical braids in R p^3 . arXiv preprint arXiv:2304.08954 , 2023

    2023 arXiv

  12. [20]

    Rubinstein

    Hyam J. Rubinstein. One-sided heegaard splittings of 3-manifolds. Pacific Journal of Mathematics , 76(1):185--200, 1978

    1978

  13. [21]

    Rubinstein

    Hyam J. Rubinstein. On 3-manifolds that have finite fundamental group and contain K lein bottles. Transactions of the American Mathematical Society , 251:129--137, 1979

    1979

  14. [22]

    Rubinstein

    Hyam J. Rubinstein. Nonorientable surfaces in some non- H aken 3-manifolds. Transactions of the American Mathematical Society , 270(2):503--524, 1982

    1982

  15. [23]

    Closed braids in 3-manifolds

    Richard Skora. Closed braids in 3-manifolds. Mathematische Zeitschrift , 211:173--187, 1992

    1992

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.