REVIEW 2 major objections 2 minor 29 references
Vop\v{e}nka's Principle, Maximum Deconstructibility, and singly-generated torsion classes
T0 review · 2 major / 2 minor · reviewed 2026-08-11 · deepseek-v4-flash
Pith's one-line read This paper proves that Vopěnka's Principle—a large-cardinal axiom—is equivalent to the assertion that every torsion class of abelian groups is generated by a single group, and to the module-theoretic principle Maximum Deconstructibility.
desk verdict The main equivalence is genuine and important; Lemma 3.3(3) is false as written, but the fix is routine and the theorem should stand. 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 mechanism is a functor from graphs to abelian groups whose hom-sets satisfy $\operatorname{Hom}_{\mathrm{Ab}}(FU,FV) \cong \mathbb{Z}^{(\operatorname{Hom}_{\mathrm{Graphs}}(U,V))}$, the free abelian group on the set of graph homomorphisms. This identity is what lets graph rigidity survive the transfer to group rigidity. The second piece is a transfinite hierarchy $T_{\alpha}^{\mathcal{C}}$ whose union is the torsion closure of a class $\mathcal{C}$, built by repeatedly taking set-indexed direct sums, extensions, and homomorphic images. The contradiction uses the fact that each of those operations preserves membership in the class ${}^{\perp_0}G$ of groups with no homomorphism to $G$, so a set-sized approximation to the torsion class can never reach a rigid group outside it. Together these pieces convert failure of Vopěnka's Principle into a torsion class that cannot be singly generated.
What would settle it
Take a pair of finite graphs $U,V$ with no graph homomorphism $U \to V$ and compute $\operatorname{Hom}_{\mathrm{Ab}}(FU,FV)$ under the paper's functor; a nonzero group homomorphism would falsify the exact homomorphism-set identity used in Theorem 3.4 and break the rigidity transfer on which the proof of statement (5) implying statement (1) depends.
Extended reading notes
Core claim
The paper's Theorem 1.1 states that Vopěnka's Principle, Maximum Deconstructibility, the deconstructibility of every filtration- and quotient-closed class of abelian groups, the deconstructibility of every torsion class of abelian groups, and the single generation of every torsion class of abelian groups are all equivalent. The previously missing direction is the last statement implying the first. Assuming Vopěnka's Principle fails, the paper builds a rigid proper class of graphs, transfers it through a graph-to-abelian-group functor with an exact homomorphism-set identity, and obtains a rigid proper class of abelian groups. Then it shows that the torsion closure of that rigid class cannot be generated by a set: a set-sized generator would appear before some stage of an ordinal-indexed torsion-closure hierarchy, leaving some rigid group outside the class, and that group's nonzero identity map would contradict rigidity. Hence the purely algebraic assertion that every torsion class is singly generated forces the reflection behavior that Vopěnka's Principle demands.
Load-bearing premise
The whole reverse direction rests on the exact homomorphism-set computation of the graph-to-abelian-group functor: if even one spurious group homomorphism appears between the images of two rigid graphs, the rigid family of groups is no longer rigid and the contradiction proving that every singly generated torsion class implies Vopěnka's Principle collapses.
Editorial extensions
If this is right
- If Vopěnka's Principle holds, then for every ring $R$ every X-Gorenstein projective class and every X-Ding projective class is deconstructible and therefore precovering, so relative homological algebra can be carried out inside those classes.
- If Vopěnka's Principle fails, there is a torsion class of abelian groups that is neither deconstructible nor singly generated; this is a concrete algebraic witness to the failure.
- The earlier theorem that Vopěnka's Principle implies every torsion class is singly generated is now an equivalence: the algebraic statement has exactly the same truth value as the large-cardinal axiom.
- Maximum Deconstructibility and Vopěnka's Principle have identical consistency strength, so any module-theoretic consequence of one is a consequence of the other.
- For torsion classes of abelian groups, deconstructibility coincides with single generation, so the existence of a small generating set is the entire content of Maximum Deconstructibility in this setting.
Reading between the lines
- The equivalence suggests a concrete way to see a Vopěnka failure: in any model where the principle fails, take the rigid sequence of graphs, push it through the exact homomorphism-set functor, and form the torsion closure; the result is a torsion class that cannot be singly generated.
- The same functor transfer may show that analogous maximum deconstructibility principles in other algebraic categories are also equivalent to Vopěnka's Principle, since the only essential input is an exact homomorphism-set embedding from a category that admits rigid proper classes.
- The open question about X-Gorenstein projective classes can be re-read: because Maximum Deconstructibility is exactly Vopěnka's Principle, proving that the full X-Gorenstein scheme implies Maximum Deconstructibility would show those homological-algebra consequences already carry large-cardinal strength.
- The proof makes Vopěnka's Principle look less like a remote set-theoretic axiom and more like a statement about the size of generating sets in accessible algebraic categories; this may make the principle testable through concrete torsion-class constructions.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper proves, in ZFC, the equivalence of Vopěnka's Principle (VP), the module-theoretic principle Maximum Deconstructibility (MD), the assertion that every class of abelian groups closed under filtrations and quotients is deconstructible, the assertion that every torsion class of abelian groups is deconstructible, and the assertion that every torsion class of abelian groups is singly-generated. The proof combines previously known implications with a new direction from single-generation of torsion classes to VP, using Przeździecki's embedding of graphs into abelian groups and the Adámek-Rosický rigid graph sequence that exists when VP fails.
Significance. If the central claim is correct, this is a significant result: it identifies a natural module-theoretic principle and a torsion-class generation principle as equivalent to a well-known large cardinal axiom, and it provides a converse to the Göbel-Shelah theorem. The paper is largely transparent, relies on external theorems in an appropriate way, and the overall strategy is elegant. However, the manuscript as written contains a technical flaw in a key lemma that is used in the proof of the new direction; the flaw is locally repairable but must be fixed before the proof is complete.
major comments (2)
- [Section 3, Lemma 3.3(3)] Lemma 3.3(3) is false as stated. The sequence ⟨T^C_α⟩ is defined using sets of representatives, so T^C_infinity is a union of representative sets and is not isomorphism-closed in general. Torsion classes are isomorphism-closed by definition, and the proof of the lemma merely asserts without justification that T^C_infinity is closed under direct sums, images, and extensions. For example, if C={Z}, then T(C) is the class of all abelian groups, while T^C_infinity is only a class of representatives for that proper class. Hence the equality T^C_infinity = T(C) cannot hold literally.
- [Section 3, proof of (5)⇒(1)] The proof of the (5)⇒(1) direction uses the false equality from Lemma 3.3(3). After assuming that a set S⊆T(G) satisfies T(S)=T(G), the paper says 'By Lemma 3.3, T(G)=T^G_infinity, so there is an ordinal γ* such that S⊆T^G_γ*'. This inference is invalid if T^G_infinity is only a class of representatives, because an arbitrary group in S need not be one of the chosen representatives. The gap is repairable: replace S by a set R of representatives, chosen from T^G_infinity, for the isomorphism classes of members of S. Since torsion classes are isomorphism-closed, T(R)=T(S)=T(G); because R is a set and R⊆T^G_infinity, one obtains R⊆T^G_γ* for some γ*, and the remainder of the argument proceeds unchanged. The manuscript should be revised to state Lemma 3.3(3) in terms of representatives and to implement this replacement explicitly.
minor comments (2)
- [Section 3, proof of (5)⇒(1)] The paper fixes 'some nonzero G ∈ G \ T^G_γ*' without justification. This is valid because the groups in a rigid proper class of abelian groups are pairwise non-isomorphic, so at most one of them is the zero group; the complement of a set in a proper class therefore contains a nonzero member. Adding a brief justification would improve clarity.
- [Section 3, Lemma 3.3] The statement of Lemma 3.3(3) should be replaced by a precise representative statement, for example that T^C_infinity contains at least one member of every isomorphism class in T(C), and the proof should be adjusted to show this weaker but sufficient property.
Circularity Check
No significant circularity: the new equivalence chain is built from external theorems, and the only author-self-cited direction is a prior published result rather than a restatement of the conclusion.
full rationale
I walked the derivation chain of Theorem 1.1. The known direction (1) implies (2) is explicitly delegated: "The implication (1) =⇒ (2) was proved in [6]." This is a self-citation to the author's own published paper, and it is load-bearing for the full equivalence statement, so I note it; but it is not circular by construction because [6] is an independently published theorem that does not assume the target equivalence. The reverse route is genuinely external: (2) implies (3) is immediate, (3) implies (4) uses standard torsion-class closure facts, (4) implies (5) is a short algebraic argument from deconstructibility to set-generation, and the main (5) implies (1) direction imports Adámek–Rosický's rigid graph sequence and Przeździecki's functor theorem. The final contradiction is structurally distinct: assuming failure of VP, the paper constructs a rigid proper class G of abelian groups and proves that T(G) is not generated by any set, so (5) fails. No fitted parameter is renamed as a prediction, no uniqueness theorem is imported from the authors' own work as a forced choice, and no ansatz is smuggled in via citation. I also considered the proof's Lemma 3.3(3), which claims T^C_infinity = T(C) even though the construction uses sets of representatives; that is a potential correctness gap about isomorphism-closure, not a circularity, so it does not raise the circularity score. Overall, the central novel claim has independent mathematical content and rests on external theorems, so I assign score 1 only for the minor self-cited forward direction.
Assumptions & free parameters
assumptions (5)
- standard math ZFC with classes is the background metatheory; statements about Vopěnka's Principle are interpreted as formal implications in ZFC.
- standard math A class of abelian groups is a torsion class iff it has the form perp0 Y for some class Y, and every torsion class is closed under filtrations (Dickson [8], Cox-Poveda-Trlifaj [7]).
- domain assumption Failure of Vopěnka's Principle gives a rigid proper class of graphs: a sequence of graphs X_alpha indexed by all ordinals with no graph homomorphism from X_alpha to X_beta for alpha not equal to beta (Adámek-Rosický [1]).
- domain assumption Przeździecki's functor F from graphs to abelian groups satisfies Hom_Ab(FU,FV) is isomorphic to the free abelian group on Hom_Graphs(U,V) for all graphs U,V (Theorem 3.4 from [25]).
- domain assumption Vopěnka's Principle implies Maximum Deconstructibility (Cox [6]).
Cite this review
Pith. "Pith review of Vop\v{e}nka's Principle, Maximum Deconstructibility, and singly-generated torsion classes." pith.science (2026). https://pith.science/paper/TLMKPTDQ
@misc{pith2026241219380,
author = {Pith},
title = {Pith review of: Vop\venka's Principle, Maximum Deconstructibility, and singly-generated torsion classes},
year = {2026},
howpublished = {\url{https://pith.science/paper/TLMKPTDQ}},
note = {Machine review of arXiv:2412.19380}
}
abstract
Deconstructibility is an often-used sufficient condition on a class $\mathcal{C}$ of modules that allows one to carry out homological algebra \emph{relative to $\mathcal{C}$}. The principle \textbf{Maximum Deconstructibility (MD)} asserts that a certain necessary condition for a class to be deconstructible is also sufficient. MD implies, for example, that the classes of Gorenstein Projective modules, Ding Projective modules, their relativized variants, and all torsion classes are deconstructible over any ring. MD was known to follow from Vop\v{e}nka's Principle and imply the existence of an $\omega_1$-strongly compact cardinal. We prove that MD is equivalent to Vop\v{e}nka's Principle, and to the assertion that each torsion class of abelian groups is generated by a single group within the class (yielding the converse of a theorem of G\"obel and Shelah).
Reference graph
Works this paper leans on
-
[1]
189, Cambridge Un iversity Press, Cambridge, 1994
Jiˇ r ´ ı Ad´ amek and Jiˇ r ´ ı Rosick´ y,Locally presentable and accessible categories , London Mathe- matical Society Lecture Note Series, vol. 189, Cambridge Un iversity Press, Cambridge, 1994
work page 1994
-
[2]
Jiˇ r ´ ı Ad´ amek and Jiˇ r ´ ı Rosick´ y,On injectivity in locally presentable categories , Trans. Amer. Math. Soc. 336 (1993), no. 2, 785–804
work page 1993
-
[3]
Asmae Ben Yassine and Jan Trlifaj, Dualizations of approximations, ℵ 1-projectivity, and Vopˇ enka’s principles, Appl. Categ. Structures 32 (2024), no. 6, Paper No. 36, 13
work page 2024
- [4]
-
[5]
Manuel Cort´ es-Izurdiaga and Jan ˇSaroch, The cotorsion pair generated by the Gorenstein projective modules and λ -pure-injective modules, arXiv preprint arXiv:2104.08602 (2023)
arXiv 2023
-
[6]
Sean Cox, Maximum deconstructibility in module categories , Journal of Pure and Applied Algebra 226 (2022), no. 5. 7
work page 2022
-
[7]
Sean Cox, Alejandro Poveda, and Jan Trlifaj, Approximation properties of torsion classes , Bull. Lond. Math. Soc. 56 (2024), no. 12, 3819–3828
work page 2024
-
[8]
Dickson, A torsion theory for Abelian categories , Trans
Spencer E. Dickson, A torsion theory for Abelian categories , Trans. Amer. Math. Soc. 121 (1966), 223–235
work page 1966
Show all 29 references
-
[9]
Nanqing Ding, Yuanlin Li, and Lixin Mao, Strongly Gorenstein flat modules , J. Aust. Math. Soc. 86 (2009), no. 3, 323–338
2009
-
[10]
Eklof and Jan Trlifaj, How to make Ext vanish , Bull
Paul C. Eklof and Jan Trlifaj, How to make Ext vanish , Bull. London Math. Soc. 33 (2001), no. 1, 41–51
2001
-
[11]
Algebra 478 (2017), 174–194
Ioannis Emmanouil, Precovers and orthogonality in the stable module category , J. Algebra 478 (2017), 174–194
2017
-
[12]
Enochs and Overtoun M
Edgar E. Enochs and Overtoun M. G. Jenda, Gorenstein injective and projective modules , Math. Z. 220 (1995), no. 4, 611–633
1995
-
[13]
30, W alter de Gruyter & Co., Berlin, 2000
, Relative homological algebra, De Gruyter Expositions in Mathematics, vol. 30, W alter de Gruyter & Co., Berlin, 2000
2000
-
[14]
Sergio Estrada, Alina Iacob, and Sinem Odabasi, Gorenstein flat and projective (pre)covers , Publ. Math. Debrecen 91 (2017), no. 1-2, 111–121
2017
-
[15]
Sergio Estrada, Alina Iacob, and Katelyn Yeomans, Gorenstein projective precovers, Mediterr. J. Math. 14 (2017), no. 1, Paper No. 33, 10
2017
-
[16]
James Gillespie, Gorenstein complexes and recollements from cotorsion pair s, Adv. Math. 291 (2016), 859–911
2016
-
[17]
, On Ding injective, Ding projective and Ding flat modules and c omplexes, Rocky Mountain J. Math. 47 (2017), no. 8, 2641–2673
2017
-
[18]
Algebra 93 (1985), no
R¨ udiger G¨ obel and Saharon Shelah, Semirigid classes of cotorsion-free abelian groups , J. Algebra 93 (1985), no. 1, 136–150
1985
-
[19]
Pure Appl
Henrik Holm, Gorenstein homological dimensions, J. Pure Appl. Algebra 189 (2004), no. 1-3, 167–193
2004
-
[20]
With a foreword by Sergio Estrada
Alina Iacob, Gorenstein homological algebra , CRC Press, Boca Raton, FL, 2019. With a foreword by Sergio Estrada
2019
-
[21]
Algebra 48 (2020), no
, Projectively coresolved Gorenstein flat and ding projectiv e modules, Comm. Algebra 48 (2020), no. 7, 2883–2893
2020
-
[22]
Thomas Jech, Set theory , Springer Monographs in Mathematics, Springer-Verlag, Be rlin,
-
[23]
Peter Jørgensen, Existence of Gorenstein projective resolutions and Tate co homology, J. Eur. Math. Soc. (JEMS) 9 (2007), no. 1, 59–76
2007
-
[24]
Daniel Murfet and Shokrollah Salarian, Totally acyclic complexes over Noetherian schemes , Adv. Math. 226 (2011), no. 2, 1096–1133
2011
-
[25]
Prze´ zdziecki, An almost full embedding of the category of graphs into the ca tegory of abelian groups , Adv
Adam J. Prze´ zdziecki, An almost full embedding of the category of graphs into the ca tegory of abelian groups , Adv. Math. 257 (2014), 527–545
2014
-
[26]
Manuel Saor ´ ın and JanˇSˇtov ´ ıˇ cek,On exact categories and applications to triangulated adjoi nts and model structures , Adv. Math. 228 (2011), no. 2, 968–1007
2011
-
[27]
Alge- bra 44 (2016), no
Jian W ang and Li Liang, A characterization of Gorenstein projective modules , Comm. Alge- bra 44 (2016), no. 4, 1420–1432
2016
-
[28]
Bin Yu, On X -Gorenstein projective dimensions and precovers , Turkish J. Math. 44 (2020), no. 5, 1768–1782. Email address : scox9@vcu.edu Department of Mathematics and Applied Mathematics, Virginia Commonwealth Uni- versity, 1015 Floyd A venue, Richmond, Virginia 23284, USA
2020
-
[2003]
The third millennium edition, revised and expanded
Reviewed August 11, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.