REVIEW 2 major objections 4 minor 21 references
Sortable simplicial complexes and their associated toric rings
T0 review · 2 major / 4 minor · reviewed 2026-08-11 · deepseek-v4-flash
Pith's one-line read The paper proves that every d-flag sortable simplicial complex is vertex decomposable, and that the associated toric and Rees algebras are Koszul, normal Cohen-Macaulay domains.
desk verdict Solid core in Sections 1–3, but Theorem 4.1 has a genuine technical gap that needs a rewrite; the stress-test counterexample, however, misreads the interval definition. 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 load-bearing object is the unit-interval simplicial complex: a pure $(d-1)$-dimensional complex in which every facet $\{i_1<\cdots<i_d\}$ has the whole interval $[i_1,i_d]$ as a clique, so its maximal cliques are intervals. The proof machinery is sortability: a set of face monomials is closed under the operation that orders the product of two monomials and alternately distributes the factors back, which yields quadratic reduced Gr\"obner bases for the toric rings and, via the $\ell$-exchange property—a condition that a needed variable can be swapped into a sorted monomial—for the Rees algebras. For the divisor class group, the key identity is the support-form description of height-one monomial primes: each prime corresponds to a supporting hyperplane of the affine semigroup cone, and the Gorenstein property is read off from the coefficients of these forms. The final vertex-decomposability theorem is carried by an induction on interval complexes that deletes a well-chosen vertex and passes to the deletion and link.
What would settle it
Check the deletion step in Theorem 4.1 on a concrete interval complex: take the 2-dimensional unit-interval complex with maximal cliques $[1,4]$ and $[3,6]$, delete vertex $4$, and see whether the resulting deletion complex is still an interval simplicial complex; if it is not, the induction as written breaks down.
Extended reading notes
Core claim
Let $\Gamma$ be a $d$-flag sortable simplicial complex on $[n]$: its minimal non-faces all have size $d$, and the multiset of face monomials is closed under the sorting operation that orders the variables of a product and alternately distributes them back. The paper proves that $\Gamma$ is sortable exactly when it is the independence complex $\operatorname{Ind}(\Delta)$ of a unit-interval simplicial complex $\Delta$—a pure complex whose every facet $\{i_1<\cdots<i_d\}$ spans an interval $[i_1,i_d]$ that is again a clique. Using this correspondence, the authors show that the toric ring $R_\Gamma=K[x_F t : F\in\Gamma]$ and the Rees algebras of the facet ideals of the pure skeletons $I(\Gamma^{[t]})$ are Koszul, normal Cohen-Macaulay domains; that all powers of $I(\Gamma^{[t]})$ have linear resolutions and satisfy the strong persistence property; that the divisor class group of $R_\Gamma$ is governed by height-one primes described by the maximal cliques of $\Delta$; and that the $a$-invariant is bounded in terms of the clique number of $\Delta$, with equality under a stated conjecture. The final theorem states that any interval simplicial complex $\Delta$ has vertex-decomposable independence complex, so any $d$-flag sortable $\Gamma$ is vertex decomposable and is Cohen-Macaulay exactly when it is pure.
Load-bearing premise
The final vertex-decomposability theorem rests on the claim that deleting a vertex from an interval simplicial complex, or passing to a link, again gives an interval simplicial complex; that preservation claim is what lets the induction in Theorem 4.1 keep going.
Editorial extensions
If this is right
- Every $d$-flag sortable simplicial complex $\Gamma$ is vertex decomposable; in particular it is Cohen-Macaulay exactly when it is pure.
- For such $\Gamma$, both $R_\Gamma$ and the Rees algebras $R(I(\Gamma^{[t]}))$ are Koszul, normal Cohen-Macaulay domains, and all powers of $I(\Gamma^{[t]})$ have linear resolutions and satisfy the strong persistence property.
- When $\Gamma$ is the independence complex of a perfect graph (the $d=2$ case), the $a$-invariant of $R_\Gamma$ equals $-\omega(G)-1$ and $R_\Gamma$ is Gorenstein if and only if all maximal cliques of $G$ have the same size.
- For $d>2$, Gorensteinness of $R_\Gamma$ forces every maximal clique of $\Delta$ to have size $2d-3$, and the $a$-invariant is bounded above by $-\lceil\omega(\Delta)/(d-1)\rceil$, with equality under Conjecture 3.3.
Reading between the lines
- Under Conjecture 3.3, the explicit height-one prime list would make the canonical module of $R_\Gamma$ computable from the maximal cliques of $\Delta$, turning Gorensteinness and the $a$-invariant into purely combinatorial checks.
- The interval structure of $\Delta$ suggests the Hilbert series of $R_\Gamma$ might be expressible directly from interval lengths; computing examples where maximal cliques overlap would test both this and the conjecture.
- If the vertex-decomposability theorem stands, it combines with standard shellability results to imply that the Stanley-Reisner rings of all interval simplicial complexes are shellable, extending the Cohen-Macaulay consequences beyond purity.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper studies d-flag sortable simplicial complexes Γ. It proves (Theorem 1.1) that such Γ are exactly independence complexes of unit-interval simplicial complexes Δ. Using this characterization and the theory of sorting orders, it shows that the Rees algebras of the facet ideals of pure skeletons are Koszul, normal Cohen-Macaulay domains (Corollary 1.6). For the toric ring R_Γ, the paper determines the a-invariant for the 2-flag (perfect graph) case (Proposition 2.1), recovers a Gorenstein characterization (Theorem 2.2), and for d>2 gives partial results on the divisor class group, canonical module, Gorenstein property, and a-invariant, some conditional on Conjecture 3.3. Finally, it claims that every interval simplicial complex has a vertex-decomposable independence complex (Theorem 4.1), yielding that every d-flag sortable complex is vertex decomposable and hence Cohen-Macaulay if and only if pure (Corollary 4.3).
Significance. If the main results hold, the paper provides a clean combinatorial characterization of Cohen-Macaulayness for d-flag sortable complexes and strong homological properties (Koszul, normal, Cohen-Macaulay) for the associated Rees and toric rings. The paper makes good use of existing machinery: it relies on established sorting orders, Gröbner basis theory, and the divisor class group descriptions from [13]. It also provides a new proof of a known Gorenstein characterization for toric rings of independence complexes of perfect graphs. The main caveat is that the vertex-decomposability theorem, which underpins the final characterization, rests on an unproved and nontrivial assertion in the proof of Theorem 4.1. Sections 1–3 are detailed and appear sound, though several statements for d>2 are conditional on Conjecture 3.3. Overall, the paper is valuable if the gap is fixed.
major comments (2)
- [Section 4, proof of Theorem 4.1] The proof asserts without proof that 'Del_Γ(i) and Lk_Γ(i) are interval simplicial complexes, as well', and this assertion is used to apply the induction hypothesis to the complexes inside Ind(·). This is load-bearing: without it, the induction step does not go through, and thus Theorem 4.1, Corollary 4.2, and Corollary 4.3 are not established. The issue is that after deleting vertex i, the vertex set is [n]\{i}, which is not an initial segment, while the definition of interval simplicial complex in the paper is made for complexes with vertex set [n]. The authors should either prove the assertion after an explicit order-preserving relabeling of the vertex set and verification of the interval condition, or supply a different argument. The surrounding text suggests that the intended objects are the complexes Φ_Del and Φ_Lk, not the independence complexes themselves, so a clarification of the notation is also needed.
- [Section 4, display before 'Notice that'] The notation is confusing: Del_Γ(i) and Lk_Γ(i) are independence complexes, yet the sentence immediately after the display says they are interval simplicial complexes; presumably the complexes inside the Ind(·) operators are meant. Please rewrite for clarity and provide the missing proof that these inner complexes are interval simplicial complexes, or replace the argument with a different induction that does not require this assertion.
minor comments (4)
- [Abstract / Introduction] The term 'd-flag sortable simplicial complex' is used without a definition in the introduction; please define it explicitly, since it is central to the paper.
- [Section 4, equation (7)] The indexing 'r_1, ..., r_n' appears to be a typo; there are m components, so the notation should be 'r_1, ..., r_m'.
- [References] Reference [12] contains a typo in the title: 'tindependence ideals' should be 'independence ideals'.
- [Proposition 3.7] The notation GΓ and Gc is easy to confuse; consider writing G^c for the complement graph.
Circularity Check
No circularity: the main derivations are self-contained or rest on published theorems, not on the results they purport to prove.
full rationale
The paper contains no fitted quantities, no prediction from fitted parameters, and no quantity defined in terms of the result it is said to derive. Theorem 1.1 is proven directly from the definitions of sortability and unit-interval complexes. Proposition 1.2 and Corollary 1.6 use external Groebner-basis theorems ([7], [9], [10], [11], [14], [15]) whose hypotheses are stated and which do not assert the paper's conclusions. Sections 2-3 use the divisor-class-group descriptions of [13] and [5] as black boxes; these are published results by overlapping authors, but invoking them is not circular because the present claims (a-invariant, Gorenstein criterion, upper bound) are derived by applying those descriptions, not by assuming the claims. The same holds for the vertex-decomposability claim: it is proved by induction in Theorem 4.1 and then applied in Corollary 4.3. The only flagged weakness is a correctness gap, not circularity: the proof of Theorem 4.1 asserts without proof that 'Del_Gamma(i) and Lk_Gamma(i) are interval simplicial complexes, as well', and that assertion is false for the unit-interval complex with maximal cliques [1,4] and [3,6] and r=2. That undermines the induction as written, but it is not a reduction of the conclusion to its own input or to a fitted parameter. Hence the circularity score is 0.
Assumptions & free parameters
assumptions (5)
- standard math Sortable monomial sets give quadratic reduced Gröbner bases for toric defining ideals ([7, Theorem 1.4]).
- standard math For a monomial ideal with the ℓ-exchange property, the Rees algebra has a quadratic reduced Gröbner basis ([10, Theorem 5.1], [4, Theorem 6.16]).
- standard math The divisor class group and canonical module of toric rings of simplicial complexes are described as in [13], and the Gorenstein criterion via support forms holds ([5, Theorem 4.3]).
- ad hoc to paper In Theorem 4.1, the complexes Del_Γ(i) and Lk_Γ(i) are interval simplicial complexes.
- domain assumption The representation of an interval simplicial complex as a union of pure skeletons of simplices whose vertex sets are intervals (equation (7)).
Cite this review
Pith. "Pith review of Sortable simplicial complexes and their associated toric rings." pith.science (2026). https://pith.science/paper/FCXFU3VV
@misc{pith2026241210113,
author = {Pith},
title = {Pith review of: Sortable simplicial complexes and their associated toric rings},
year = {2026},
howpublished = {\url{https://pith.science/paper/FCXFU3VV}},
note = {Machine review of arXiv:2412.10113}
}
abstract
Let $\Gamma$ be a $d$-flag sortable simplicial complex. We consider the toric ring $R_{\Gamma}=K[{\bf x}_Ft:F\in \Gamma]$ and the Rees algebra of the facet ideals $I(\Gamma^{[i]})$ of pure skeletons of $\Gamma$. We show that these algebras are Koszul, normal Cohen-Macaulay domains. Moreover, we study the Gorenstein property, the canonical module, and the $a$-invariant of the normal domain $R_{\Gamma}$ by investigating its divisor class group. Finally, it is shown that any $d$-flag sortable simplicial complex is vertex decomposable, which provides a characterization of the Cohen-Macaulay property of such complexes.
Reference graph
Works this paper leans on
-
[13]
Toric rings attached to simplicial complexes
J. Herzog, S. Moradi, A. Asloob Qureshi, Toric rings attached t o simplicial complexes (2023), arXiv:2302.03653
work page Pith review arXiv 2023
-
[1]
A. Almousa, K. VandeBogert, Determinantal facet ideals for sm aller minors, Arch. Math. 118(3) (2022) 247–256
work page 2022
-
[2]
B. Benedetti, L. Seccia, M. Varbaro, Hamiltonian paths, unit-int erval complexes, and deter- minantal facet ideals, Adv. in Appl. Math. 141(2022), Paper No. 102407, 55
work page 2022
- [3]
-
[4]
V. Ene, J. Herzog, Gr¨ obner bases in commutative algebra, Ame rican Mathematical Soc., 2011
work page 2011
-
[5]
The toric ring of one dimensional simplicial complexes
A. Ficarra, J. Herzog, D. Stamate, The toric ring of one dimensio nal simplicial complexes, 2024, to appear in Rocky Mountain J. Math., available at arxiv.org/abs/2306.05020
work page Pith review arXiv 2024
- [6]
- [7]
Show all 21 references
-
[8]
Herzog, T
J. Herzog, T. Hibi, S. Moradi, A. Asloob Qureshi, The divisor class g roup of a discrete polymatroid, 2024, J. Comb. Theory Ser. A. 205, 105869
2024
-
[9]
Herzog, T
J. Herzog, T. Hibi, H. Ohsugi, Binomial ideals, Graduate Texts in Ma thematics. Springer, New York, 2018
2018
-
[10]
Herzog, T
J. Herzog, T. Hibi, M. Vladoiu. Ideals of fiber type and polymatro ids, Osaka J. Math. 42 (2005) 807-829
2005
-
[11]
Herzog, T
J. Herzog, T. Hibi, X. Zheng, Monomial ideals whose powers have a linear resolution, Math. Scand. 95 (2004), no. 1, 23–32
2004
-
[12]
Herzog, F
J. Herzog, F. Khosh-Ahang, S. Moradi, M. Rahimbeigi, Sortable simplicial complexes and tindependence ideals of proper interval graphs, Electron. J. Com b. 27(1) (2020) P1.65
2020
-
[14]
Herzog, A
J. Herzog, A. Asloob Qureshi, Persistence and stability proper ties of powers of ideals, J. Pure Appl. Algebra 219 (2015), 530-542
2015
-
[15]
Hochster, Rings of invariants of tori, Cohen-Macaulay rings generated by monomials, and polytopes, Ann
M. Hochster, Rings of invariants of tori, Cohen-Macaulay rings generated by monomials, and polytopes, Ann. Math., 96 (1972), 228–235
1972
-
[16]
Moradi, Normal Rees algebras arising from vertex decompos able simplicial complexes, arXiv:2311.15135
S. Moradi, Normal Rees algebras arising from vertex decompos able simplicial complexes, arXiv:2311.15135. 16
-
[17]
Moradi, F
S. Moradi, F. Khosh-Ahang, On vertex decomposable simplicial c omplexes and their Alexan- der duals, Math. Scand. 118 (2016), no. 1, 43–56
2016
-
[18]
Ohsugi, T
H. Ohsugi, T. Hibi, Convex polytopes all of whose reverse lexicog raphic initial ideals are squarefree, Proc. Amer. Math. Soc. 129 (2001), No. 9, 2541–2546
2001
-
[19]
Ohsugi, T
H. Ohsugi, T. Hibi, Special simplices and Gorenstein toric rings, J. Combin. Theory Ser. A 113 (2006), no. 4, 718–725
2006
-
[20]
Sturmfels, Gr¨ obner Bases and Convex Polytopes, vol.8, American Mathematical Soc., 1996
B. Sturmfels, Gr¨ obner Bases and Convex Polytopes, vol.8, American Mathematical Soc., 1996
1996
-
[21]
Woodroofe, Vertex decomposable graphs and obstruction s to shellability, Proc
R. Woodroofe, Vertex decomposable graphs and obstruction s to shellability, Proc. Amer. Math. Soc. 137 (2009) 3235–3246. Antonino Ficarra, Departamento de Matem ´atica, Escola de Ci ˆ encias e Tecnolo- gia, Centro de Investigac ¸˜ao, Matem´atica e Aplicac ¸˜oes, Instituto de I...
2009
Reviewed August 11, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.