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 →
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 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.
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
- 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.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
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)
- [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.
- [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)
- [Section 1, introduction] 'Besides of crossing change' should be 'Besides crossing change'; also 'an local operation' should be 'a local operation'.
- [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.
- [Throughout] Expressions like 'm2−m 2' should be typeset as (m^2-m)/2 to avoid ambiguity about the intended parentheses.
- [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.
- [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.
- [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
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
assumptions (4)
- domain assumption Theorem 2.4: for orientable surface Sigma_g, rank(M_L) = r - n - 1 + rank(N_L)
- domain assumption Crosscap addition disjoint from L does not affect r, n, rank(M_L), or rank(N_L)
- ad hoc to paper The half twist preserves the number of components and the rank of the homology matrix
- ad hoc to paper The block matrices M13, M23, M41, M43 have the forms displayed
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.
Forward citations
Cited by 1 Pith paper
-
When is arc crossing change an unknotting operation?
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
-
[4]
Jiawei Cheng, Zhiyun Cheng, Jinwen Xu, and Jieyao Zheng, Region crossing change on surfaces , Mathematische Zeitschrift 300 (2022), no. 3, 2289–2308
work page 2022
-
[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
work page 2012
-
[2]
Haruko Aida, Unknotting operation for Polygonal type, Tokyo J. Math. 15 (1992), no. 1, 111–121
work page 1992
-
[3]
J. W. Alexander, Topological invariants of knots and links , Trans. Amer. Math. Soc. 30 (1928), no. 2, 275–306
work page 1928
-
[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
2012
-
[6]
Zhiyun Cheng, When is region crossing change an unknotting operation? , Math. Proc. Cambridge Philos. Soc 155 (2013), no. 2, 257–269
work page 2013
-
[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
work page 2020
-
[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
work page 2018
Show all 17 references
-
[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
2015
-
[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
2016
-
[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
2021
-
[12]
Hitoshi Murakami, Some metrics on classical knots, Math. Ann. 270 (1985), 35–45
1985
-
[13]
Hitoshi Murakami and Yasutaka Nakanishi, On a certain move generating link-homology, Math. Ann. 284 (1989), 75–89
1989
-
[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
2024
-
[15]
4-5, 693–704
Martin Scharlemann, Crossing changes, Chaos Solitons Fractals 9 (1998), no. 4-5, 693–704
1998
-
[16]
Ayaka Shimizu, Region crossing change is an unknotting operation, J. Math. Soc. Japan66 (2014), no. 3, 693–708
2014
-
[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...
2020
Reviewed August 7, 2026 · model on record in the stance chip above.
Discussion (0). Sign in to comment.