{"id":"70836cf9-5eb2-4754-bd10-2c92b2645d81","arxiv_id":"1908.05826","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For fiber-type and hypersolvable arrangements, the holonomy Lie algebra is an iterated almost-direct product of free Lie algebras with ranks equal to the arrangement exponents.","lead":"This paper shows that the algebraic invariant called the holonomy Lie algebra has a simple step-by-step structure for fiber-type and hypersolvable hyperplane arrangements. The result gives the full internal structure, not just dimension counts, and recovers known formulas as a byproduct.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The free-factor claim in Theorem 3.13 rests entirely on Lemma 3.14, whose proof is a reference to an unstated part of Jambu 1990; if that injectivity fails, the kernel is a proper quotient of L(A_h) and the almost-direct product structure collapses.","rationale":"The reader and I identify the same critical point: Theorem 3.13 depends on an injectivity statement that is not proved. I found no internal contradiction in the rest of the proof: Prop 3.4's snake lemma argument is standard, the closedness of A_v in A is correctly argued, and the d=0 case is genuinely trivial. The only mismatch is that Lemma 3.14's stated hypotheses are not met in d=0, but this is an expository gap, not a mathematical failure. The real question is whether the cited Part 2 of Jambu's proof supplies the assertion in the exact generality needed. Because the paper gives no details, the central claim is conditionally supported at best. This does not change the reader's verdict: CONDITIONAL remains appropriate. The test I propose (recovering the referenced argument or a degree-by-degree computation on a concrete fiber-type arrangement) would settle the point.","tokens_in":73,"tokens_out":9579,"duration_ms":500517,"concrete_test":"Obtain [Jam90, Thm 4.3.1], write out Part 2 of its proof in the notation of §2-3 of the present paper, and check that: (i) the hypotheses match Lemma 3.14 for B=A_v in a strictly linearly fibered arrangement, and (ii) no additional coordinate condition is silently used. As a complementary computational check, compute I(A)∩L(A_h) in degrees 2 and 3 for the rank-3 braid arrangement of Example 3.9 using the generators given in §2.2; a nonzero element would refute Lemma 3.14, while a zero intersection in those degrees supports the claimed injectivity.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central theorem (Thm 3.13) reduces the structure of h(A) to the claim, in Lemma 3.14, that for B⊂A with ⋂B modular of rank r(A)-1, the free subalgebra L(A\\B) meets the defining ideal I(A) trivially. This is the step that converts Ker π* from the quotient L(A\\B)/(L(A\\B)∩I(A)), obtained in Prop 3.4, into the free Lie algebra L(A\\B). The proof of Lemma 3.14 is only the sentence 'This is exactly Part 2 of the proof of [Jam90] Theorem 4.3.1.' No statement of that theorem part, no verification of hypothesis match, and no d=0 discussion are given. In the strictly linearly fibered setting, Prop 3.12 gives the modular-rank condition only when d>0; for d=0, r(⟨K⟩)=r(A), so the lemma's rank hypothesis is not satisfied. That case is trivial because A_h is empty, but the proof of Thm 3.13 does not say so. If Lemma 3.14 is false, the kernel is not the free Lie algebra generated by A_h, and Corollaries 3.15 and 4.8 lose their content. The paper is plausible and the result is likely true, but the load-bearing step is not established within the manuscript.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies the holonomy Lie algebra h(A) of central hyperplane arrangements, using Kohno's combinatorial description in terms of rank-two flats. For a strictly linearly fibered arrangement, the authors prove that h(A) is an almost-direct product of h(A_v), the holonomy Lie algebra of the vertical subarrangement, and the free Lie algebra on the horizontal hyperplanes (Theorem 3.13). Iterating this result yields the paper's main structural claim: for a fiber-type arrangement with exponents (d_1,...,d_ell), h(A) is an iterated almost-direct product of free Lie algebras L(d_1),...,L(d_ell) (Corollary 3.15). Using Jambu-Papadima's deformation method, the same conclusion is extended to hypersolvable arrangements (Corollary 4.8). The paper is concise and mostly clearly written, and it explicitly connects the results to earlier LCS formulas.","tokens_in":9938,"tokens_out":8009,"duration_ms":75589,"significance":"If the main theorem is fully established, the result is significant: it describes the entire Lie algebra structure of h(A) for fiber-type and hypersolvable arrangements, not merely the graded dimensions of the lower central series. This is a natural Lie-algebra analogue of the Falk-Randell almost-direct product decomposition for fundamental groups, and it strengthens earlier LCS formulas of Kohno, Falk-Randell, and Jambu-Papadima. The paper's approach via closed subarrangements is a reasonable extension of the work of Lima-Filho and Schenck, and the use of Kohno's theorem and Jambu-Papadima deformation theory is legitimate. However, the central structural conclusion rests on Lemma 3.14, whose proof in the manuscript is only a reference to part of a proof in Jambu's paper; as written, the main theorem is therefore conditional on an external result whose exact statement and hypothesis verification are not supplied.","major_comments":[{"comment":"The proof of Lemma 3.14 consists solely of the sentence 'This is exactly Part 2 of the proof of [Jam90] Theorem 4.3.1.' This lemma is load-bearing: it is exactly what turns the kernel of pi_* from the quotient L(A\\B)/(L(A\\B) \\cap I(A)) obtained in Proposition 3.4 into the free Lie algebra L(A\\B). Without this injectivity, the almost-direct product structure in Theorem 3.13 and the free factors in Corollaries 3.15 and 4.8 collapse. The manuscript neither states the content of 'Part 2' nor verifies that the hypotheses of Jambu's proof are satisfied in the present setting, nor does it discuss the case d=0. Please provide a self-contained proof of Lemma 3.14, or, if the result is truly Jambu's, state the relevant theorem in full and explicitly check every hypothesis.","section":"Section 3.2, Lemma 3.14"},{"comment":"The proof of Theorem 3.13 says that the conditions of Lemma 3.14 are satisfied 'in view of Proposition 3.12.' However, Proposition 3.12 gives the required rank condition r(\\langle K\\rangle)=r(A)-1 only when d>0; when d=0, it gives r(\\langle K\\rangle)=r(A), so Lemma 3.14 does not apply. The statement is still true in the d=0 case because A_h is empty and L(A_h)=0, but the proof as written contains a gap. Please add an explicit sentence treating the d=0 case.","section":"Section 3.2, proof of Theorem 3.13"},{"comment":"The extension to hypersolvable arrangements depends on Theorem 4.6 and Corollary 4.7, which are cited rather than proved. This is acceptable if the cited results are correct, but the proof should explicitly explain why the equality L_{\\le 2}(A)=L_{\\le 2}(\\tilde A(1)) from Theorem 4.6 implies an isomorphism of holonomy Lie algebras h(A) \\cong h(\\tilde A(1)). This fact is true because the defining ideal of h(A) is generated by rank-two flat data, but it is not stated, and it is the bridge that carries the almost-direct product structure from the fiber-type case to the hypersolvable case.","section":"Section 4, Corollary 4.8"}],"minor_comments":[{"comment":"The phrase 'almost-directed product' should be 'almost-direct product' throughout the paper.","section":"Section 3.1, Proposition 3.4"},{"comment":"The symbol L(A) is used both for the intersection lattice and for the free Lie algebra on A. This overloading is likely to confuse readers; please use separate notation, such as \\mathcal{L}(A) for the lattice and L(A) for the free Lie algebra.","section":"Section 2, Notation"},{"comment":"The statement that h(A) 'has a graded vector space decomposition as a direct sum of free Lie algebras' is potentially misleading because the decomposition is not a Lie algebra direct sum. Please explicitly say 'as graded vector spaces' to avoid suggesting a Lie algebra isomorphism.","section":"Section 3.2, Theorem 3.11"},{"comment":"In the proof there is a typo: 'A if of fiber-type' should be 'A is of fiber-type'.","section":"Corollary 3.16"},{"comment":"The reference list heading 'Reference' should be 'References'.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper is honest about its dependence on Jambu's work, but in its current form it outsources a load-bearing lemma to 'Part 2 of the proof' of a theorem, which is not a self-contained proof. I would encourage the editor to require the authors to provide a proof of Lemma 3.14, or at minimum to state Jambu's result in full and verify all hypotheses explicitly. I found no indication of circularity, and the claimed result is plausible; the issue is completeness rather than correctness."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"What you should know: this paper proves that holonomy Lie algebras of fiber-type (supersolvable) arrangements are iterated almost-direct products of free Lie algebras with ranks the exponents, and extends to hypersolvable arrangements via Jambu-Papadima deformation. That is a genuinely new structural result—Jambu 1990 only had the vector-space decomposition, and Falk-Randell got the group-level version. The authors get the Lie algebra extension structure, which is not in the literature.\n\nThe proof strategy is clean. For a closed subarrangement B ⊂ A, they show h(A) is an almost-direct product of h(B) and the kernel of the projection, which is normally L(A\\B)/(L(A\\B) ∩ I(A)). Then Theorem 3.13 has to identify that kernel with a free Lie algebra L(A_h), and that is where the paper becomes thin: Lemma 3.14 asserts L(A\\B) ∩ I(A) = 0 for certain modular B, and its proof is one sentence: 'This is exactly Part 2 of the proof of [Jam90] Theorem 4.3.1.' No statement of that part, no verification of the hypothesis match, no handling of the d = 0 case where the rank hypothesis is not met. The stress-test concern is accurate; if that lemma is false, the kernel collapses, and Corollaries 3.15 and 4.8 lose their content. I think the lemma is likely true and the result correct—this is not a circular step, just an unproved external fact—but the paper should do its own lifting here, especially since the whole theorem depends on it.\n\nMinor: the arXiv metadata title says 'Holonomy Lie algebra of a geometric lattice,' which matches the abstract text about lattices and oriented matroids. The actual full text is titled 'Holonomy Lie algebra of a fiber-type arrangement' and has no geometric-lattice content. That will not help anyone citing this paper, and the authors should sort out which version is intended. The proofs of Lemma 3.2 and Proposition 3.4 generalizing Lima-Filho-Schenck are well done. The LCS formula corollaries are known results, honestly presented as byproducts.\n\nBottom line: a solid, useful structural theorem with one load-bearing reference standing in for a proof. Let a referee decide on the lemma; if it is accepted, the paper is a worthwhile contribution to arrangement theory. Send it out.","headline":"New Lie-algebra structure theorem for fiber-type and hypersolvable arrangements; the main proof leans on an unproved 1990 lemma and the arXiv title doesn't match the text.","tokens_in":10502,"tokens_out":2537,"would_cite":true,"duration_ms":21680,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["52C35","17B70","17B01","05B35"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that the holonomy Lie algebra of a fiber-type arrangement is an iterated almost-direct product of free Lie algebras whose ranks are the exponents, and that the same structure holds for hypersolvable arrangements.","keywords":["holonomy Lie algebra","hyperplane arrangements","fiber-type arrangements","supersolvable arrangements","hypersolvable arrangements","almost-direct product","free Lie algebra","lower central series"],"falsifier":"Take a fiber-type arrangement $\\mathcal{A}$ and the horizontal set $\\mathcal{A}_h$ from its top fibration, and compute in degree 3 the kernel of the inclusion $L(\\mathcal{A}_h)\\to h(\\mathcal{A})=L(\\mathcal{A})/I(\\mathcal{A})$. A nonzero kernel element, such as a linear combination of brackets $[x,[y,z]]$ with $x,y,z\\in\\mathcal{A}_h$ that lies in $I(\\mathcal{A})$, would violate Lemma 3.14 and collapse the iterated almost-direct product theorem; for the braid arrangement $\\mathcal{A}_4$ with exponents $(1,0,2,3)$, this check is a finite linear algebra computation from the rank-two flats.","tokens_in":9418,"feed_emoji":"📐","tokens_out":16766,"duration_ms":154024,"temperature":0.7,"pith_summary":"The paper proves that the holonomy Lie algebra of a fiber-type hyperplane arrangement—a quotient of the free Lie algebra on the hyperplanes by relations attached to rank-two flats—splits as an iterated almost-direct product of free Lie algebras whose ranks are exactly the arrangement's exponents. The same splitting is proved for hypersolvable arrangements, a wider class, by deforming them to fiber-type arrangements without changing rank-two intersections. Since this Lie algebra is isomorphic to the rational associated graded Lie algebra of the fundamental group of the complement, all lower-central-series ranks follow, together with the known LCS formula. The argument is combinatorial, working with subarrangement pairs rather than with the topology of fibrations.","feed_headline":"Fiber-type holonomy Lie algebras split into free Lie factors","feed_subtitle":"The result fixes the entire rational Lie algebra of the complement, not just its graded dimensions.","key_machinery":"The carrying object is the defining ideal $I(\\mathcal{A})$ of the holonomy Lie algebra: $I(\\mathcal{A})$ is generated by all brackets $[H,\\Sigma_L]$, where $L$ is a rank-two flat and $\\Sigma_L$ is the sum of the hyperplanes below $L$, and $h(\\mathcal{A})=L(\\mathcal{A})/I(\\mathcal{A})$. The proof's engine is the closed-subarrangement pair: for a closed subarrangement $\\mathcal{B}\\subset\\mathcal{A}$, contraction onto $\\mathcal{B}$ gives a split surjection $h(\\mathcal{A})\\to h(\\mathcal{B})$; a rewriting lemma moves any bracket touching $\\mathcal{A}\\setminus\\mathcal{B}$ into brackets built only from the removed hyperplanes, so the kernel is generated by those hyperplanes. Lemma 3.14, whose proof is imported from an existing theorem, then shows the kernel is free, provided the common intersection of $\\mathcal{B}$ is a modular element of rank $r(\\mathcal{A})-1$. Stacking one such split at each level of the fibration tower yields iterated almost-direct products of free Lie algebras.","core_discovery":"For a strictly linearly fibered arrangement $\\mathcal{A}$, split the hyperplanes into vertical ones $\\mathcal{A}_v$ (containing the kernel of the projection) and horizontal ones $\\mathcal{A}_h$. The paper shows that $\\mathcal{A}_v$ is closed in $\\mathcal{A}$ and that the projection $h(\\mathcal{A})\\to h(\\mathcal{A}_v)$ has kernel generated by $\\mathcal{A}_h$; Lemma 3.14 guarantees this kernel is literally the free Lie algebra $L(\\mathcal{A}_h)$. Hence $h(\\mathcal{A})$ is an almost-direct product of $h(\\mathcal{A}_v)$ and $L(\\mathcal{A}_h)$, and iterating along the tower of fibrations gives the central theorem: for a fiber-type arrangement with exponents $(d_1,\\dots,d_\\ell)$, the holonomy Lie algebra is an iterated almost-direct product of $L(d_1),\\dots,L(d_\\ell)$. A vertical deformation that preserves rank-two intersections transfers the same conclusion to hypersolvable arrangements, with $d_i=\\#\\mathcal{A}_i-\\#\\mathcal{A}_{i-1}$ along a composition series; in particular $\\dim h(\\mathcal{A})_j=\\sum_i\\dim L(d_i)_j$, and the LCS formula follows.","pith_inferences":["The same closed-subarrangement mechanism should yield an almost-direct product for solvable pairs of geometric lattices, as the abstract announces; the body carries out the arrangement case, leaving the full lattice statement to the announced framework.","A self-contained proof of the imported injectivity lemma would likely generalize the theorem to any pair whose removed hyperplanes have a modular intersection of rank $r(\\mathcal{A})-1$, without requiring an actual linear fibration; triangle-complete pairs of graphic arrangements are a natural place to test this.","Two hypersolvable arrangements with the same exponents have the same graded Lie algebra dimensions and LCS series, but the theorem does not claim the extension data of the almost-direct product is determined by exponents; comparing arrangements with equal exponents and different rank-two lattices would show whether the isomorphism type is finer than the exponent data."],"forward_implications":["The entire rational associated graded Lie algebra of $\\pi_1(M(\\mathcal{A}))$ is determined for fiber-type arrangements, not just its Hilbert series.","For every $j$, $\\phi_j(\\mathcal{A})=\\sum_{i=1}^{\\ell}\\dim L(d_i)_j$, making all lower-central-series ranks explicit.","The LCS formula $\\prod_{j\\ge1}(1-t^j)^{\\phi_j(\\mathcal{A})}=\\prod_{i=1}^{\\ell}(1-d_i t)$ follows directly from the Lie algebra structure.","Hypersolvable arrangements, including all supersolvable ones, inherit the same iterated almost-direct product structure via the rank-two-preserving vertical deformation.","Since holonomy Lie algebras depend only on the intersection lattice up to rank two, any arrangement with the same $L_{\\le2}$ as a fiber-type arrangement has an isomorphic holonomy Lie algebra."],"supporting_citations":[{"why":"It identifies the holonomy Lie algebra of the complement with the rational associated graded Lie algebra of the fundamental group, so structural theorems about $h(\\mathcal{A})$ control lower-central-series ranks.","marker":"[Koh83]"},{"why":"It proves fiber-type arrangements are exactly the supersolvable ones and supplies the modular element of rank $r(\\mathcal{A})-1$ used in Lemma 3.14.","marker":"[Ter86]"},{"why":"It contains the theorem to which Lemma 3.14 is delegated; its injectivity statement $L(\\mathcal{A}\\setminus\\mathcal{B})\\cap I(\\mathcal{A})=0$ is the load-bearing step for freeness of the kernel.","marker":"[Jam90]"},{"why":"It defines closed and solvable subarrangements and hypersolvable arrangements, and provides the split-extension formalism that Proposition 3.4 generalizes.","marker":"[JP98]"},{"why":"It provides the vertical deformation that preserves $L_{\\le2}$ and converts hypersolvable arrangements into fiber-type arrangements, transferring the main theorem.","marker":"[JP02]"},{"why":"It defines fiber-type arrangements, proves the group-level iterated almost-direct product of free groups, and establishes the LCS formula that the Lie algebra result mirrors.","marker":"[FR85]"},{"why":"It supplies the closed-subarrangement technique for rewriting brackets through removed hyperplanes that the proof generalizes from graphic arrangements to all closed pairs.","marker":"[LFS09]"}],"fun_headline_variants":["Fiber-type holonomy splits into free Lie factors","Holonomy Lie algebra: iterated free product for fiber-type","Supersolvable arrangements: holonomy is free Lie tower","Holonomy of lattice: almost-direct product with free Lie","Free Lie algebras decompose hypersolvable holonomy"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof needs the claim that a subarrangement whose common intersection is a modular element of rank one less than the whole arrangement contributes no accidental relations, so the removed hyperplanes generate a genuinely free subalgebra; this claim is imported from an existing proof rather than demonstrated here.","fun_headline_variants_meta":{"raw":{"variants":["Fiber-type holonomy splits into free Lie factors","Holonomy Lie algebra: iterated free product for fiber-type","Supersolvable arrangements: holonomy is free Lie tower","Holonomy of lattice: almost-direct product with free Lie","Free Lie algebras decompose hypersolvable holonomy"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00034,"raw_usage":{"total_tokens":1856,"prompt_tokens":909,"completion_tokens":947,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":525,"completion_tokens_details":{"reasoning_tokens":867}},"tokens_in":525,"tokens_out":947,"duration_ms":8679,"temperature":1.0,"reasoning_tokens":867,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:04:45.464538+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a fiber-type arrangement $\\mathcal{A}$ and the horizontal set $\\mathcal{A}_h$ from its top fibration, and compute in degree 3 the kernel of the inclusion $L(\\mathcal{A}_h)\\to h(\\mathcal{A})=L(\\mathcal{A})/I(\\mathcal{A})$. A nonzero kernel element, such as a linear combination of brackets $[x,[y,z]]$ with $x,y,z\\in\\mathcal{A}_h$ that lies in $I(\\mathcal{A})$, would violate Lemma 3.14 and collapse the iterated almost-direct product theorem; for the braid arrangement $\\mathcal{A}_4$ with exponents $(1,0,2,3)$, this check is a finite linear algebra computation from the rank-two flats.","supporting_citations":[],"review_version":1}