Pith. sign in

REVIEW 2 major objections 6 minor 1 cited by

Region crossing change on nonorientable surfaces

T0 review · 2 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

Pith's one-line read This paper proves a rank formula that classifies link diagrams on nonorientable surfaces up to region crossing changes.

desk verdict Plausible and likely true, but the proof of the main theorem has a load-bearing unproved assertion about the half twist preserving components and homology rank. read the letter →

arxiv 2506.05885 v1 pith:Q4QBL3DD submitted 2025-06-06 math.GT

classification math.GT MSC 57K1057M15
keywords regioncrossingchangenonorientablesurfaceincidencematrixhomologylinkdiagrambicoloringKleinbottleprojectiveplane
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 classifies link diagrams drawn on a nonorientable closed surface, such as a projective plane or Klein bottle, up to region crossing change: the operation that switches every crossing on the boundary of a chosen region. Its main theorem gives a formula for the rank of the incidence matrix $M_L$ in terms of the diagram's regions, components, and the homology classes of those components. If correct, the formula implies that among the $2^c$ diagrams sharing one shadow, exactly $2^{c - \operatorname{rank}(M_L)}$ are inequivalent under region crossing changes. This completes the orientable case and settles the last open case in the classification of diagrams on closed surfaces.

What carries the argument

The central objects are the $\mathbb{Z}_2$ incidence matrix $M_L$, whose rows record which regions border each crossing an odd number of times, and the homology matrix $N_L$, whose rows record the homology classes of the link components in $H_1(N_g;\mathbb{Z}_2)$. The proof's mechanism is a half twist along an arc that converts $N_{2g}$ into $gT^2$; a block-matrix computation shows the twist adds $(m^2-m)/2$ to the rank of $M_L$, matching the added crossings and regions, after which the orientable theorem applies. The bicoloring criterion of Theorem 4.2 then turns the rank computation into a checkable condition for any chosen set of crossings.

What would settle it

Take a link diagram on the Klein bottle with two arcs crossing the cutting edge $a_g$ that belong to different components, perform the half twist, and count components and homology rank before and after. If the half twist merges the two components or changes the rank of $N_L$, the comparison $\operatorname{rank}(M_{L''}) = \operatorname{rank}(M_{L'}) + (m^2-m)/2$ would fail, contradicting Theorem 1.2; one could also compute both sides directly for such an example.

Watch

Extended reading notes

Core claim

Theorem 1.2 states that for a link diagram $L = K_1 \cup \dots \cup K_n$ on $N_g$, the connected sum of $g$ real projective planes, $\operatorname{rank}(M_L) = r - n - 1 + \operatorname{rank}(N_L)$ over $\mathbb{Z}_2$, where $M_L$ is the region-crossing incidence matrix and $N_L$ records the $\mathbb{Z}_2$-homology classes of the components. Together with Theorem 4.2, which characterizes admissible sets of crossings by the existence of a bicoloring whose 1-colored semi-arcs form a null-homologous cycle, this gives a complete description of when two diagrams on the same shadow differ by region crossing changes. The proof reduces the nonorientable surface to an orientable one by a half twist along an edge of a polygonal presentation, applies the orientable theorem, and computes the effect of the twist on the incidence matrix.

Load-bearing premise

The proof assumes without proof that the half twist along the edge $a_g$ preserves both the number of link components and the rank of the homology matrix; if that assertion fails for some diagram, the reduction to the orientable case collapses.

Editorial extensions

If this is right

  • For a fixed shadow with $c$ crossings on $N_g$, the $2^c$ crossing assignments split into exactly $2^{c-r+n+1-\operatorname{rank}(N_L)}$ classes under region crossing changes.
  • A chosen set of crossings is region crossing change admissible exactly when the corresponding linear system over $\mathbb{Z}_2$ has a solution, equivalently when the bicoloring condition of Theorem 4.2 holds.
  • The incidence matrix of a link diagram on $N_g$ has full column rank precisely when $r - n - 1 + \operatorname{rank}(N_L) = c$, so region crossing change is an unknotting operation exactly in that case.
  • The orientable formula of Theorem 2.4 appears as the special case where the diagram lies on an orientable surface or on the orientable part of a nonorientable one.

Reading between the lines

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

  • A testable next step is computing both sides of the half-twist comparison for small diagrams on the Klein bottle; this would either certify the stated invariance of component number and homology rank or expose a counterexample.
  • Because the paper uses the double counting rule, the classification does not automatically transfer to Shimizu's original single counting convention; translating it would require tracking how many times a region touches a crossing.
  • The same incidence-matrix framework likely extends to link diagrams on surfaces with boundary or to spatial graph diagrams on nonorientable surfaces, where the admissibility question would depend on a similar homology condition.
  • The bicoloring characterization suggests a direct algorithm for finding the minimal set of region crossing changes realizing a chosen crossing set, by searching for a null-homologous color cycle in the semi-arc graph.
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

2 major / 6 minor

Summary. The paper studies region crossing changes on link diagrams on closed nonorientable surfaces N_g. The main theorem (Theorem 1.2) asserts that for a link diagram L with n components on N_g, the rank of the Z_2 incidence matrix M_L equals r - n - 1 + rank(N_L), where r is the number of regions and N_L is the Z_2 homology matrix of the components. The proof reduces the nonorientable case to the orientable theorem of [4] by cutting along a one-sided curve and applying a half twist to the arcs meeting it, then computing the rank of the resulting incidence matrix by block elimination. The paper also gives a criterion (Theorem 4.2) for a set of crossings to be region crossing change admissible via a null-homologous bicoloring. The classification follows by counting c - rank(M_L) equivalence classes.

Significance. If the proof gap identified below is repaired, the result would be a complete and natural extension of the orientable classification in [4] to nonorientable surfaces, with the same formal structure. The paper includes a worked example that verifies the formula on T^2 and the Klein bottle. The bicoloring characterization is a useful diagrammatic tool. The main technical contribution is the reduction of the nonorientable case to the orientable one, but the key invariance step in that reduction is not proved.

major comments (2)
  1. [Section 3 (proof of Theorem 1.2)] The sentence 'Note that the half twist preserves the number of components and the rank of homology matrix' is asserted without proof or citation. This is load-bearing: Theorem 2.4 is applied to L'' on gT^2 with the same n and the same rank(N), so if the half twist changes the number of components or the homology map, the conclusion does not follow. A half twist along a one-sided curve is a cut-and-paste operation that changes the pairing between the two copies of the cut edge; for two arcs from different components that both cross a_g, reversing the gluing can merge them into a single component. The R2 moves described before the half twist do not alter the intersection set of the diagram with a_g, so they do not place the arcs in a normal form that makes the claim automatic. The authors need a lemma that defines the half twist precisely and proves that n and rank(N) are unchanged, or that reduces to a normal form where this is evident.
  2. [Section 3, block form of M_{L''}] The matrices M13, M23, M41, M43 are exhibited without derivation, and the displayed M13 is ambiguous (the expression '1 01×(m−3) 1 1 ...' mixes scalar entries and zero-row blocks without parentheses). The claimed rank equality rank(M_{L''}) = rank(M_{L'}) + (m^2-m)/2 depends on these block forms and on the specific column eliminations converting M43 to an identity matrix. Because Figure 6 is the sole source of these blocks, the authors should include an explicit construction of the blocks from the half-twisted diagram, with all dimensions checked, and a step-by-step account of the column operations.
minor comments (6)
  1. [Section 1, introduction] 'Besides of crossing change' should be 'Besides crossing change'; also 'an local operation' should be 'a local operation'.
  2. [Section 2, Example 2.3] The notation 'on 2RP2' should be 'on N_2' for consistency with the definition of N_g in the introduction.
  3. [Throughout] Expressions like 'm2−m 2' should be typeset as (m^2-m)/2 to avoid ambiguity about the intended parentheses.
  4. [Section 3, Lemma 3.1] The proof of Lemma 3.1 sketches a rank computation and says 'the proof above still holds' when some of R'_2,...,R'_5 coincide; this should be written out explicitly, since the row operations are not identical in that case.
  5. [Section 4, Theorem 4.2] In the 'if' direction of Theorem 4.2, the union of 1-colored semi-arcs may have self-intersections at crossings not in P where both strands are colored 1; the proof should indicate how to pass from such a 1-chain to an embedded subsurface, for example by smoothing the vertices and taking a regular neighborhood.
  6. [Theorem 2.4 and references] Reference [4] is a paper coauthored by the first author of the present paper; the dependence on it is appropriate, but Theorem 2.4 should be clearly presented as an external result rather than as part of the current derivation.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: Theorem 1.2 is reduced to the externally published orientable formula of [4] plus self-contained linear algebra; the unproved half-twist invariance claim is a proof gap, not a circular input.

full rationale

The derivation of Theorem 1.2 is not circular. The nonorientable case is reduced to Theorem 2.4, the orientable-surface rank formula, quoted from the published paper [4] (Mathematische Zeitschrift 300, 2022). Although [4] shares the present first author, Theorem 2.4 is an external, peer-reviewed result whose stated assumptions (closed orientable surface) do not include the target result (closed nonorientable surface), so by the independence standard for citations it does not raise the circularity score. The remaining steps are self-contained: Lemma 3.1 verifies that r − rank(ML) is invariant under the second Reidemeister move by an explicit rank computation; the block-matrix argument in Section 3 establishes rank(ML'') = rank(ML') + (m^2 − m)/2 via elementary column operations on the explicitly displayed matrices M41, M43, M13, and M23; and Theorem 4.2 is proven directly, constructing the subsurface S whose boundary is the 1-colored semi-arcs in one direction and coloring regions black/white in the other. No parameter is fitted to data and no quantity is defined in terms of the quantity it predicts. The one load-bearing unproved assertion, 'Note that the half twist preserves the number of components and the rank of homology matrix' (Section 3, proof of Theorem 1.2), is a proof gap rather than circularity: it does not make the theorem's conclusion an input, and the invariance genuinely needs a separate geometric lemma, since the reversal gluing along a_g can in principle alter component counts. That concern belongs under correctness risk, not circularity, so the score is 0.

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

No free parameters or invented entities. The proof rests on two external structural facts (orientable rank formula and homology matrix behavior) and two unproved, paper-specific geometric claims: the crosscap-addition invariance and the half-twist preservation of n and rank(N). These claims are the real burden of the proof.

assumptions (4)
  • domain assumption Theorem 2.4: for orientable surface Sigma_g, rank(M_L) = r - n - 1 + rank(N_L)
    Published in [4] and used as the base case; it is an external theorem, but the proof of Theorem 1.2 depends on it without rederiving it.
  • domain assumption Crosscap addition disjoint from L does not affect r, n, rank(M_L), or rank(N_L)
    Used in the odd-to-even reduction in Section 3 without proof. Plausible because the new crosscap lies in a region and adds no crossings, but the region count is not explicitly argued.
  • ad hoc to paper The half twist preserves the number of components and the rank of the homology matrix
    Stated without proof in Section 3. Load-bearing because the orientable formula is applied to the twisted diagram with unchanged n and rank(N).
  • ad hoc to paper The block matrices M13, M23, M41, M43 have the forms displayed
    These identities are asserted from Figure 6 and the preceding Reidemeister moves; they are not derived algebraically and the column operations that transform M43 to the identity are not shown in detail.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Region crossing change on nonorientable surfaces." pith.science (2026). https://pith.science/paper/Q4QBL3DD

@misc{pith2026250605885,
  author       = {Pith},
  title        = {Pith review of: Region crossing change on nonorientable surfaces},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Q4QBL3DD}},
  note         = {Machine review of arXiv:2506.05885}
}
read the original abstract

In this paper, we give a classification of link diagrams on nonorientable surfaces up to region crossing changes.

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. When is arc crossing change an unknotting operation?

    math.GT 2025-09 conditional novelty 7.0 of 10

    Type I arc crossing change unknots every link diagram; type II leaves a Hopf link residue when the total linking number is odd, and any two crossings in an alternating knot diagram can be switched.

Reference graph

Works this paper leans on

17 extracted references · 16 canonical work pages · cited by 1 Pith paper

  1. [4]

    3, 2289–2308

    Jiawei Cheng, Zhiyun Cheng, Jinwen Xu, and Jieyao Zheng, Region crossing change on surfaces , Mathematische Zeitschrift 300 (2022), no. 3, 2289–2308

  2. [1]

    Knot Theory Ramifications 21 (2012), no

    Kazushi Ahara and Masaaki Suzuki, An integral region choice problem on knot projection , J. Knot Theory Ramifications 21 (2012), no. 11, 1250119, 20 pp

  3. [2]

    Haruko Aida, Unknotting operation for Polygonal type, Tokyo J. Math. 15 (1992), no. 1, 111–121

  4. [3]

    J. W. Alexander, Topological invariants of knots and links , Trans. Amer. Math. Soc. 30 (1928), no. 2, 275–306

  5. [5]

    7, 1487–1495

    Zhiyun Cheng and Hongzhu Gao, On region crossing change and incidence matrix , Science China Mathematics 55 (2012), no. 7, 1487–1495

  6. [6]

    Zhiyun Cheng, When is region crossing change an unknotting operation? , Math. Proc. Cambridge Philos. Soc 155 (2013), no. 2, 257–269

  7. [7]

    Russell,Ineffective sets and the region crossing change operation, J

    Miles Clikeman, Rachel Morris, and Heather M. Russell,Ineffective sets and the region crossing change operation, J. Knot Theory Ramifications 29 (2020), no. 3, 2050010, 17 pp

  8. [8]

    Russell, Equivalence of edge bicolored graphs on surfaces, Electron

    Oliver Dasbach and Heather M. Russell, Equivalence of edge bicolored graphs on surfaces, Electron. J. Combin. 25 (2018), no. 1, Paper 1.59, 15 pp

Show all 17 references
  1. [9]

    Knot Theory Ramifications 24 (2015), no

    Kenta Hayano, Ayaka Shimizu, and Reiko Shinjo, Region crossing change on spatial-graph diagrams, J. Knot Theory Ramifications 24 (2015), no. 8, 1550045, 12 pp

  2. [10]

    Knot Theory Ramifications 25 (2016), no

    Ayumu Inoue and Ryo Shimizu, A subspecies of region crossing change, region freeze crossing change , J. Knot Theory Ramifications 25 (2016), no. 14, 1650075, 9 pp

  3. [11]

    Knot Theory Ram- ifications 30 (2021), no

    Kengo Kawamura, Region crossing change, bicolored diagram and Arf invariant, J. Knot Theory Ram- ifications 30 (2021), no. 5, 2150029, 17 pp

  4. [12]

    Hitoshi Murakami, Some metrics on classical knots, Math. Ann. 270 (1985), 35–45

  5. [13]

    Hitoshi Murakami and Yasutaka Nakanishi, On a certain move generating link-homology, Math. Ann. 284 (1989), 75–89

  6. [14]

    Knot Theory Ramifications 33 (2024), no

    Tokio Oshikiri, Ayaka Shimizu, and Junya Tamura, Region crossing change on origami and link , J. Knot Theory Ramifications 33 (2024), no. 7, 2450019, 14 pp

  7. [15]

    4-5, 693–704

    Martin Scharlemann, Crossing changes, Chaos Solitons Fractals 9 (1998), no. 4-5, 693–704

  8. [16]

    Ayaka Shimizu, Region crossing change is an unknotting operation, J. Math. Soc. Japan66 (2014), no. 3, 693–708

  9. [17]

    Knot Theory Ramifications 29 (2020), no

    Ayaka Shimizu and Rinno Takahashi, Region crossing change on spatial theta-curves, J. Knot Theory Ramifications 29 (2020), no. 5, 2050028, 11 pp. SCHOOL OF MATHEMATICAL SCIENCES , L ABORATORY OF MATHEMATICS AND COMPLEX SYS- TEMS , MOE, B EIJING NORMAL UNIVERSITY , B EIJING , 1...

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.