REVIEW 3 major objections 4 minor 9 references
Band of topological groups
T0 review · 3 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash
Pith's one-line read A band of topological groups is metrizable exactly when every H-class is, and its quotients by full normal subcryptogroups are Hausdorff exactly when the subcryptogroup is closed.
desk verdict Genuine but uneven extension of the semilattice results to bands; main theorems are right but two proofs have repairable gaps and several results are deferred to the authors' earlier paper. 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 carrying object is the $\mathcal{H}$-class decomposition of a cryptogroup: $S$ is a band of groups with each $\mathcal{H}$-class $G_\alpha$ a clopen topological group and the topology generated by the union of the group topologies. The proof of metrizability uses the star operation $(UV)^* = \bigcup\{(uV)^*\}$, where $(uV)^*=\{x : u^{-1}x\in V,\ u_0=x_0\}$, to transfer neighbourhood bases between idempotents and arbitrary points; the metric $d(x,y)=\min\{1,d_\alpha(x,y)\}$ within one $\mathcal{H}$-class and $1$ between different classes then reproduces the topology. For quotients, the central object is the congruence $\rho_N$ defined by $a\,\rho_N\, b$ iff $a^{-1}b\in N$ and $a_0=b_0$, whose classes are written $(xN)^*$; Theorem 6.8 compares closedness of $\rho_N$, closedness of $N$, and Hausdorffness of $S/N$ through the natural quotient map.
What would settle it
To test Theorem 6.8, take a concrete band of topological groups (for instance the mod-10 example in the paper), pick a full normal subcryptogroup $N$, and compute whether $N$ is closed, whether $\rho_N$ is closed in $S\times S$, and whether $S/N$ is Hausdorff; the theorem predicts the same answer for all three, so one mismatch refutes it. To test Theorem 4.13, check directly that the metric $d(x,y)=\min\{1,d_\alpha(x,y)\}$ inside one $\mathcal{H}$-class and $1$ between different classes generates the original topology; the theorem says it always does because each $\mathcal{H}$-class is clopen.
Extended reading notes
Core claim
The central claim is Theorem 4.13: a band of topological groups $(S,\tau)$ is metrizable if and only if each $\mathcal{H}$-class is metrizable. The proof is constructive: given metrics $d_\alpha$ on the H-classes $G_\alpha$, form bounded metrics and define a global metric $d$ by $d(x,y)=\bar d_\alpha(x,y)$ when $x,y$ lie in the same $\mathcal{H}$-class $G_\alpha$, and $d(x,y)=1$ otherwise. The open balls of $d$ are open in $\tau$ and generate it because every $\mathcal{H}$-class is clopen. The paper's second main claim is Theorem 6.8: for a full normal subcryptogroup $N$ of $S$, the quotient $S/N$ is Hausdorff if and only if the congruence $\rho_N$ is closed in $S\times S$ if and only if $N$ is closed in $S$. Along the way the paper shows Hausdorff bands are completely regular and locally compact Hausdorff bands are normal, and that first countability, separability, and second countability reduce to H-class conditions, with $E(S)$ countable in the latter two cases.
Load-bearing premise
The Hausdorff-quotient theorem depends on the unproved assumption that $S/N$ is again a band of topological groups and that the quotient map is open; if either fails for a noncommutative band, the claimed equivalence can fail.
Editorial extensions
If this is right
- Every Hausdorff band of topological groups is completely regular, and every locally compact Hausdorff band is normal, so separation functions exist for closed sets and points.
- A Hausdorff band of topological groups is metrizable exactly when it is first countable, because the classical first-countability criterion for topological groups applies to each H-class.
- When $E(S)$ is countable, the whole band is separable or second countable exactly when each H-class is, so countability of the band of idempotents transfers fiberwise properties.
- Quotienting by a full normal subcryptogroup always yields another band of topological groups; if the subcryptogroup is closed, the quotient is Hausdorff.
- Discrete full subcryptogroups of Hausdorff countably compact bands are finite, and of Hausdorff Lindelöf bands they are countable.
Reading between the lines
- If every H-class admits a complete metric, the constructed metric is also complete, so complete metrizability reduces to the H-classes, although the paper does not state this.
- The distance-1-between-classes metric makes a band of topological groups a metric sum of its group fibers, which opens the way to coarse-geometric or large-scale questions about such bands.
- The closedness criterion for $\rho_N$ suggests that non-closed full normal subcryptogroups are exactly the source of non-Hausdorff quotients, and closure of such a subcryptogroup yields a Hausdorff quotient by the same machinery.
- Theorem 6.7 is deferred to a semilattice analogue without a proof in this text; a direct proof for noncommutative bands would make the Hausdorff-quotient theorem self-contained.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper studies bands of topological groups, i.e., cryptogroups whose Green H-classes are open topological groups. It characterizes when a topological cryptogroup is a band of topological groups (Theorem 2.5), constructs a canonical topology from local neighbourhood systems at idempotents (Theorem 3.7), proves that a band of topological groups is metrizable if and only if each H-class is metrizable (Theorem 4.13), and proves that for a full normal subcryptogroup N, the quotient S/N is Hausdorff if and only if the congruence ρ_N is closed in S × S, which in turn is equivalent to N being closed in S (Theorem 6.8). Many supporting results are either stated without proof or deferred to the authors' earlier paper on semilattices of topological groups [5].
Significance. If correct, the metrizability criterion and the Hausdorff quotient criterion are clean structural results: each H-class is a clopen topological group, so a global metric can be assembled from bounded metrics on the clopen summands, and quotient Hausdorffness reduces to group-quotient separation inside the H-classes. The paper also provides explicit examples and a substantial construction theorem (Theorem 3.7) with no free parameters and no circular reasoning. However, full verification is currently impeded by missing proofs for the transfer from the semilattice case to the noncommutative band case, and by a flawed line in the proof of Theorem 4.13. The gaps appear repairable, and the main claims are plausible, but they require additional work before the paper can be accepted.
major comments (3)
- [§4, Theorem 4.13] The displayed equality G_beta = union_{n>=1} B(y,1/n) is false when G_beta contains a point z with d_beta(y,z) = 1, because any ball B(y,1/n) in S has radius less than or equal to 1 and therefore excludes such z. The intended conclusion that G_beta is open in the metric topology is nevertheless true, since B(y,1/2) is contained in G_beta; the proof should be corrected accordingly.
- [§6, Theorem 6.7 and Lemma 6.6] Theorem 6.7, which asserts that S/N is a band of topological groups, is not proved in the paper; the text merely says 'Similar to [5, Theorem 5.10]', even though [5] treats semilattices rather than arbitrary bands. Lemma 6.6 is also stated without proof. This is load-bearing because Theorem 6.8 uses S/N being a band of topological groups and uses Lemma 6.6 to identify H-classes under ψ.
- [§6, proof of Theorem 6.8(iii)⇒(i)] The proof asserts that the natural mapping ψ: S → S/N is open and uses this to conclude that ψ(H_u), ψ(H_v), ψ(U_2), and ψ(V_2) are open in S/N. No proof of openness is supplied. Openness is not automatic for a quotient map; the paper should justify it, for example by showing that ψ^{-1}(ψ(U)) ∩ H_e = (U ∩ H_e)(N ∩ H_e) is open in the topological group H_e for each idempotent e. Without this, the separation argument in (iii)⇒(i) is incomplete.
minor comments (4)
- [Abstract] The phrase 'if and, ρ_N is closed in S × S' should read 'if and only if ρ_N is closed in S × S'.
- [§3, proof of Theorem 3.7, Step 2] In the sentence 'Now, we show that (U_1)* ⊆ W', the intended set is (U_1 y)*, not (U_1)*; the missing y appears in the next line but the notation should be fixed.
- [§5, Theorem 5.7] The proof says that discreteness of K in itself and Hausdorffness of S imply that K is open in its closure; this implication is not automatic and needs a justification, for example by proving that each discrete subgroup K ∩ H_e is closed in the topological group H_e.
- [§4 and §5] Theorems 4.1, 4.7, and 5.6 are stated with only a reference to [5] and no proof; since [5] treats semilattices rather than bands, the authors should either include the proofs or state explicitly why the arguments carry over verbatim.
Circularity Check
No circular reduction found; the main metric and quotient arguments are proved from the definitions, though Theorem 6.7 is deferred to the authors' earlier [5] and is a verification gap, not a circularity.
full rationale
The central derivations are not circular. Theorem 4.13 proves metrizability by explicitly defining a metric from bounded metrics on each H-class, with distance 1 between distinct H-classes, and then verifying that the metric topology equals the given topology using the defining base property of a band of topological groups. No fitted parameter or assumption of the conclusion is used. Theorem 2.5 and Theorem 3.7 are genuine equivalence proofs, not definitions of the target result in terms of itself. The only recurring device is reference to the authors' earlier paper [5], with phrases such as 'Similar to the proof of [5, Theorem 3.1]' and 'Similar to the [5, Theorem 5.10]'. These are self-citations, but they are to a published source with analogous proofs rather than to the conclusion of the present paper. The most load-bearing such instance is Theorem 6.7, whose proof is not written out and is instead attributed to [5, Theorem 5.10]; Theorem 6.8(iii)⇒(i) then relies on the natural map ψ being open, also asserted without proof. This is a genuine correctness and verification risk, but it is not circular reasoning: the Hausdorff equivalence is not assumed as an input, the quoted theorem does not contain the desired conclusion as a premise, and the construction does not define the target into existence. Hence no circular step is present; the score reflects the nontrivial reliance on self-citation and deferred proofs rather than any reduction by construction.
Assumptions & free parameters
assumptions (6)
- standard math Each H-class of a cryptogroup is a group and H is a congruence.
- standard math The map x maps to x0 is continuous in a topological cryptogroup.
- standard math A Hausdorff topological group is metrizable if and only if it is first countable.
- standard math Every locally compact Hausdorff topological group is normal.
- standard math Character of a point is preserved under dense subspaces in regular spaces.
- standard math If f: X x Y to Z is continuous, A and B are compact, and W is an open set containing f(A x B), then there are open U, V with A subset U, B subset V, and f(U x V) subset W.
Cite this review
Pith. "Pith review of Band of topological groups." pith.science (2026). https://pith.science/paper/EJXB2K7J
@misc{pith2026250605956,
author = {Pith},
title = {Pith review of: Band of topological groups},
year = {2026},
howpublished = {\url{https://pith.science/paper/EJXB2K7J}},
note = {Machine review of arXiv:2506.05956}
}
abstract
In this article, we construct a band of topological groups from a cryptogroup. Also, we prove that a band of topological groups is metrizable if and only if each $\mathcal{H}$-class is metrizable. Finally, we demonstrate that if $S$ is a band of topological groups and $N$ is a full normal subcryptogroup of $S$, then $S/N$ is Hausdorff if and only if $\rho_{_N}$ is closed in $S \times S$ if and, $\rho_{_N}$ is closed in $S \times S$ if and only if $N$ is closed in $S$.
Reference graph
Works this paper leans on
-
[5]
https://doi.org/10.1080/00927872.2021.1909606
Maity, S.K., Paul, M.,Semilattice of topological groups, Communications in Algebra 49 (2021), 3905–3925. https://doi.org/10.1080/00927872.2021.1909606
arXiv 2021
-
[1]
M.,An introduction to semigroup theory, United Kingdom: Academic Press, 1976
Howie, J. M.,An introduction to semigroup theory, United Kingdom: Academic Press, 1976
work page 1976
-
[2]
R.,Completely Regular Semigroups, United Kingdom: Wiley, 1999
Petrich, M., Reilly, N. R.,Completely Regular Semigroups, United Kingdom: Wiley, 1999
work page 1999
-
[3]
Munkres, J.,Topology, United Kingdom: Pearson, 2017
work page 2017
-
[4]
Carruth, J. H., Hildebrant, J. A., Koch, R. J.,The Theory of Topological Semigroups, Switzerland: M. Dekker, 1983
work page 1983
-
[6]
Husain, T.,Introduction to Topological Groups, United States: Dover Publications, 2018
work page 2018
-
[7]
Birkhoff, G.,A note on topological groups, Compositio Math. V ol. 3 (1936), 427 - 430
work page 1936
-
[8]
V., Tkachenko, M.,Topological Groups and Related Structures, Netherlands: Atlantis Press,2008
Arkhangel’ski ˘i, A. V., Tkachenko, M.,Topological Groups and Related Structures, Netherlands: Atlantis Press,2008
work page 2008
Show all 9 references
-
[9]
Engelking, R.,General Topology, Poland: PWN, 1977. 14
1977
Reviewed August 7, 2026 · model on record in the stance chip above.
Discussion (0). Sign in to comment.