Spectral sequences, Massey products and homology of covering spaces

Basic Information

• Paper ID: 2511.11893
• Title: Spectral sequences, Massey products and homology of covering spaces
• Authors: Yongqiang Liu, Laurentiu Maxim, Botong Wang
• Classification: math.AT (Algebraic Topology)
• Submission Date: November 18, 2025 (arXiv submission)
• Paper Link: https://arxiv.org/abs/2511.11893

Abstract

This paper revisits the equivariant spectral sequence considered by Papadima-Suciu and proves that all its differentials can be computed via higher-order Massey products. As a first application, the authors generalize Pajitnov's relationship between the Jordan block sizes of the eigenvalue 1 part of the Alexander module and the lengths of nonzero Massey products in cohomology to arbitrary field coefficients. The paper also provides computable upper bounds for the mod p Betti numbers of prime power cyclic covers and the ranks of cohomology groups of rank-one local systems over fields of prime order. Under appropriate conditions, these bounds improve upon results of Papadima-Suciu. The results are further specialized to hyperplane arrangement complements, proving that the vanishing of higher-order Massey products implies that the mod p Betti numbers of prime p-tower cyclic covers are determined by the combinatorial structure.

Research Background and Motivation

Problem Background

1. Core Problem: Study the relationship between the homological properties of covering spaces and the cohomological algebraic structure of the base space, particularly Massey products.
2. Classical Theory: For infinite cyclic covers $X_\nu$ (determined by a group homomorphism $\nu: \pi_1(X) \twoheadrightarrow \mathbb{Z}$), the Alexander module $H_i(X_\nu, K)$ is a finitely generated $R = K[t^{\pm 1}]$-module. When $\nu$ is induced by a fibration $f: X \to S^1$, extensive work has studied the relationship between Jordan block sizes of the monodromy action on $H_*(F,\mathbb{C})$ and Massey products of $X$.
3. Existing Results:
   • Fernández-Gray-Morgan: Relationship between nonzero length-2 Massey products and Jordan blocks of size greater than 1
   • Papadima-Suciu: The case of 1-formal spaces, where the $H^1(F,\mathbb{C})$ part of eigenvalue 1 is semisimple
   • Bazzoni-Fernández-Muñoz: Jordan blocks of size 2 imply nonzero triple Massey products
   • Pajitnov: Identified the relationship between the lengths of specific nonzero higher-order Massey products and Jordan block sizes of the eigenvalue 1 part (complex coefficients)

Research Motivation

Core Motivation: Generalize Pajitnov's results to field coefficients of positive characteristic. This generalization is crucial because:

1. Failure of Formality: Compact Kähler manifolds and hyperplane arrangement complements are typically not formal over $\mathbb{F}_p$ (despite being formal over $\mathbb{C}$)
2. Theoretical Gap: Existing spectral sequence theory for the torsion part of Alexander modules primarily focuses on complex coefficients
3. Combinatorial Invariance: For hyperplane arrangements, understanding Massey products in positive characteristic is essential for determining the combinatorial invariance of Betti numbers of covering spaces

Core Contributions

The main contributions of this paper include:

1. Complete Characterization of Spectral Sequences and Massey Products: Proves that all differentials of the Papadima-Suciu equivariant spectral sequence can be computed via higher-order Massey products (Proposition 2.2)
2. Generalization of Pajitnov's Theorem: Extends the relationship between Jordan block sizes and Massey product lengths to arbitrary field coefficients (Theorem 1.1)
3. Jordan Block Bounds for Algebraic Varieties: Provides explicit upper bounds for Jordan block sizes for complex algebraic varieties satisfying weight conditions (Corollary 1.2)
4. Betti Number Bounds for Prime Power Cyclic Covers: Establishes computable upper bounds for mod p Betti numbers of $p^r$-fold covers $X_r$ (Proposition 1.3): $b_i(X_r, \mathbb{F}_p) \leq b_i(X, \mathbb{F}_p) + (p^r - 1) \cdot \beta_i(X, \eta_p)$
5. Improved Bounds for Rank-One Local Systems: Under the condition of no p-torsion, provides finer bounds than Papadima-Suciu (Theorem 1.5)
6. Combinatorial Determinacy for Hyperplane Arrangements: Proves that when higher-order Massey products vanish, the mod p Betti numbers of prime p-tower cyclic covers are determined by the combinatorial structure

Detailed Methodology

Core Technical Framework

The paper's core technical approach establishes a "Massey-type spectral sequence" and proves its duality with the Papadima-Suciu J-adic spectral sequence.

Cochain Algebra Characterization of Alexander Modules

Setup: Let $X$ be a connected finite CW complex, $\nu: \pi_1(X) \twoheadrightarrow \mathbb{Z}$ a group surjection. Fix a field $K$ and denote $R = K[\mathbb{Z}] \cong K[t^{\pm 1}]$.

Key Construction (Section 2):

1. Local System: Consider the local system $L_\nu$ on $X$ with fiber $R$, where the fundamental group representation is induced by $\nu$. There is an isomorphism: $H_i(X, L_\nu) \cong H_i(X_\nu, K)$
2. Twisted Chain Complex: For any $m \in \mathbb{Z}_{>0}$, define the twisted chain complex of $R/(s^{m+1})$-modules (where $s = t-1$): $C_*(X, \eta, m) = (C_*(X) \otimes_K R/(s^{m+1}), \partial + s\eta)$ with boundary map: $(\partial + s\eta)(\Delta) = \partial\Delta + s(\Delta \cap \eta)$
3. Key Isomorphism (Proposition 2.2): Proves that $(\partial + s\eta)^2 = 0$ and establishes the isomorphism: $H_i(X, L_\nu \otimes_R R/(s^{m+1})) \cong H_i(C_*(X, \eta, m))$

Proof Technique: Through explicit construction of simplicial complexes, using the Cartesian diagram:

X_ν ----f'_ν----> R
 |                |
 p                q
 |                |
 v                v
 X  -----f_ν----> S¹

constructs a chain complex isomorphism $\Phi$ and verifies its compatibility with boundary maps.

Spectral Sequence Construction

Bicomplex Structure (Section 3):

View the cochain complex $(C^*(X,K) \otimes_K \widehat{R}, \delta \otimes \text{id} + (\eta \cup -) \otimes s)$ as a bicomplex: $A^{i,j} := C^{i+j}(X,K) \otimes_K s^i \cong C^{i+j}(X,K)$

• Vertical maps: $\delta \otimes \text{id}$
• Horizontal maps: cup product with $\eta$

Pages of the Spectral Sequence:

1. $E_1$-page: $E_1^{i,j} = H^{i+j}(X,K)$ with differential $d_1^{i,j}(\alpha) = [\eta] \cup \alpha$
2. $E_2$-page: $E_2^{i,j} = H^{i+j}(H^*(X,K), [\eta] \cup -)$ which is the Aomoto cohomology, with dimension defined as the Aomoto Betti number: $\beta_i(X, \eta_p) := \dim_{\mathbb{F}_p} H^i(H^*(X,\mathbb{F}_p), [\eta_p] \cup -)$
3. Higher Page Differentials: For $k \geq 2$, the differential $d_k$ is computed by $(k+1)$-fold Massey products

Definition of Massey Products

Definition 3.4: For $\omega \in H^i(X,K)$ and $k \geq 1$, the $(k+1)$-fold Massey product $\langle[\eta], \omega\rangle_k$ is defined as:

Given a defining system $\{\alpha_1, \ldots, \alpha_k\}$ satisfying:

• $\delta\alpha_1 = 0, [\alpha_1] = \omega$
• $\delta\alpha_2 = \eta \cup \alpha_1$
• $\vdots$
• $\delta\alpha_k = \eta \cup \alpha_{k-1}$

the $(k+1)$-fold Massey product is the cohomology class of $\eta \cup \alpha_k$ modulo indeterminacy.

Key Properties:

• When $k=2$, this recovers the classical triple Massey product $\langle[\eta], [\eta], \omega\rangle$
• The spectral sequence differential $d_k$ precisely computes these $(k+1)$-fold Massey products

Main Theorems and Proof Strategies

Theorem 1.1 (Characterization of Jordan Block Sizes)

Statement: For $i \geq 0$, the maximum Jordan block size in the eigenvalue 1 part of $H_i(X_\nu, K)$ is one less than the length of the highest nonzero Massey product in $i$-th cohomology related to $\eta$. In particular, if all higher-order Massey products are trivial, then the eigenvalue 1 part is semisimple.

Proof Strategy:

1. Use the fact that $\widehat{R} = K[[s]]$ is a principal ideal domain
2. Any bounded finitely generated $\widehat{R}$-module complex is quasi-isomorphic to a direct sum of complexes of the form $0 \to \widehat{R} \to 0$ or $0 \to \widehat{R} \xrightarrow{\cdot s^j} \widehat{R} \to 0$
3. The spectral sequence of the second type of complex degenerates at the $(j+1)$-th page, while the module $\widehat{R}/(s^j)$ has Jordan block size $j$
4. Therefore, the maximum Jordan block size = $\min_k\{d_k^{p,q} = 0 \text{ for all } p+q=i\} - 1$
5. By the Massey product interpretation of spectral sequence differentials, this is precisely one less than the highest nonzero Massey product length

Corollary 1.2 (Application to Algebraic Varieties)

Statement: Let $X$ be an $n$-dimensional complex algebraic variety (possibly singular), assume $W_0H^1(X,\mathbb{C}) = 0$ (weight filtration). Then for any $\nu: \pi_1(X) \twoheadrightarrow \mathbb{Z}$:

• The Jordan block size of the eigenvalue 1 part of $H_i(X_\nu, \mathbb{C})$ is $\leq \min\{2i+2, 2n\}$
• For $H^1$, the bound can be improved to 3
• If additionally $W_1H^1(X,\mathbb{C}) = 0$, the bound is $\min\{i+1, n\}$

Key Points of Proof: Use the fact that the Sullivan de Rham cdga of complex algebraic varieties carries a weight filtration with differentials preserving weights. The assumption $W_0H^1 = 0$ implies that the weight of $d_k(\alpha)$ increases by at least $k$, so $d_k(\alpha) = 0$ when $k > \min\{2i+2, 2n\}$.

Propositions 1.3 and 1.4 (Betti Numbers of Prime Power Covers)

Proposition 1.3: For a $p^r$-fold cover $X_r \to X$, we have the inequality: $b_i(X_r, \mathbb{F}_p) \leq b_i(X, \mathbb{F}_p) + (p^r - 1) \cdot \beta_i(X, \eta_p)$

Proof: Use the truncated spectral sequence (with $p^r$ columns), estimating from the $E_2$-page: $b_i(X_r, \mathbb{F}_p) \leq \sum_{k=i-p^r+1}^{i} \dim E_2^{i-k,k}$ Expanding and using the structure of the $E_2$-page yields the desired inequality.

Proposition 1.4: For $p^r > 2$, equality holds if and only if the spectral sequence degenerates at the $E_2$-page. In particular, if there exists a nontrivial $k$-fold Massey product ($3 \leq k \leq p^r$), then the inequality in degree $i$ is strict.

Theorem 1.5 (Improved Bounds for Rank-One Local Systems)

Statement: Let $\lambda \in \mathbb{C}^*$ be a primitive $p$-th root of unity, and $L_\lambda$ the corresponding rank-one local system. Then: $b_i(X, L_\lambda) \leq \beta_i(X, \eta_p) + \frac{b_i(X,\mathbb{F}_p) - b_i(X,\mathbb{C})}{p-1}$

If $H^*(X,\mathbb{Z})$ has no $p$-torsion, then: $b_i(X, L_\lambda) \leq \beta_i(X, \eta_p)$

If there exists a nontrivial $k$-fold Massey product ($3 \leq k \leq p$), then the inequality is strict.

Proof Strategy:

1. Consider the $p$-fold cover $Y \to X$, using $b_i(Y,\mathbb{C}) = b_i(X,\mathbb{C}) + (p-1)b_i(X,L_\lambda)$
2. By the universal coefficient theorem and Proposition 1.3: $b_i(Y,\mathbb{C}) \leq b_i(Y,\mathbb{F}_p) \leq b_i(X,\mathbb{F}_p) + (p-1)\beta_i(X,\eta_p)$
3. Combining yields the desired inequality
4. When there is no $p$-torsion, $b_i(Y,\mathbb{C}) = b_i(Y,\mathbb{F}_p)$, giving the improved bound

Applications to Hyperplane Arrangements

Formality Problem

Background: Hyperplane arrangement complements $X$ are formal over $\mathbb{C}$ (in the sense of Sullivan rational homotopy theory), but typically not formal over $\mathbb{F}_p$.

Matei Example (Example 5.1): The arrangement $A(p,1,3)$ in $\mathbb{C}^3$ is defined by: $z_1 \cdot z_2 \cdot z_3 \cdot \prod_{1 \leq i < j \leq 3}(z_i^p - z_j^p)$

For $p > 2$, Matei proved that its complement $X$ has a nontrivial triple Massey product on $H^2(X,\mathbb{F}_p)$.

Numerical Verification: For $p=3$, computing via Fox calculus yields: $b_1(X, L_\lambda) = 1 < \beta_1(X, \eta_3) = 2$ consistent with the strict inequality of Theorem 1.5.

Application of Cirici-Horel Theorem

Theorem 5.2 (Cirici-Horel): Let $X$ be a hyperplane arrangement complement with arrangement defined over an $\ell$-adic field $K$ ($\ell \neq p$), with residue field of cardinality $q = \ell^m$. Let $h$ be the order of $q$ in $\mathbb{F}_p^*$. If $(k-2)/h \notin \mathbb{Z}$, then all $k$-fold Massey products in $H^*(X,\mathbb{F}_p)$ are trivial.

Corollary: For graphic arrangements, taking $\ell = q = 2$:

• When $p = 3$, $h = 2$, all $k$-fold Massey products with $k \neq 2$ are trivial
• When $p = 5$, $h = 4$, all $k$-fold Massey products with $k \not\equiv 2 \pmod{4}$ are trivial

Therefore, by Proposition 1.4, the Betti numbers of the $p$-fold cover $Y$ are: $b_i(Y,\mathbb{F}_p) = b_i(X,\mathbb{F}_p) + (p-1) \cdot \beta_i(X,\eta_p)$ completely determined by combinatorial data.

Technical Innovations

1. Duality Observation

Innovation: The isomorphism established in Proposition 2.2 reveals a duality relationship between the Massey-type spectral sequence and the Papadima-Suciu J-adic spectral sequence. This observation is key because:

• The cohomological version is better suited for handling cup products and Massey products
• The chain complex version is better suited for understanding the homology of covering spaces

2. Explicit Construction

Through explicit construction of simplicial complexes (rather than abstract categorical arguments), the paper provides:

• Geometric meaning of the twisted boundary operator $\partial + s\eta$
• Concrete computational methods for spectral sequence differentials
• Clear relationship with the action of covering transformations

3. Technical Handling of Positive Characteristic Generalization

Key technical challenges:

• Sign Conventions: Detailed treatment of signs in cap and cup products (Remark 2.1), particularly important in positive characteristic
• Leibniz Rule: Verifying $(\partial + s\eta)^2 = 0$ requires careful sign tracking
• Simplicial Structure: Using the special simplicial structure of $S^1$ to simplify proofs

4. Refinement of Bounds

Compared to the general bound of Papadima-Suciu $b_i(X,L_\lambda) \leq \beta_i(X,\eta_p)$ (for all $p^r$-th roots of unity $\lambda$), this paper:

• Provides explicit bounds for prime power covers (Proposition 1.3)
• Recovers the Papadima-Suciu bound under no-torsion assumptions (Theorem 1.5)
• Proves that nontrivial Massey products lead to strict inequalities (Propositions 1.4, Theorem 1.5)

Related Work

Historical Context

1. Mapping Class Theory:
   • Fernández-Gray-Morgan (1991): Massey products and Kähler manifolds
   • Papadima-Suciu (2010): 1-formality and semisimplicity
   • Bazzoni-Fernández-Muñoz (2015): Jordan blocks and triple Massey products
2. Alexander Module Theory:
   • Papadima-Suciu (2010): J-adic spectral sequence
   • Pajitnov (2017, 2019): Massey products and Jordan blocks (complex coefficients)
   • Budur-Liu-Wang (2018): Monodromy theorems for Kähler manifolds and quasi-projective varieties
3. Hyperplane Arrangements:
   • Matei (2006): Non-formality examples in positive characteristic
   • Yoshinaga (2020): Combinatorial determinacy of double covers
   • Cirici-Horel (2022): Étale cohomology and formality

Positioning of This Paper

Relative to Pajitnov:

• Generalizes to arbitrary field coefficients (especially positive characteristic)
• Uses spectral sequence methods rather than Novikov homology

Relative to Papadima-Suciu:

• Explicitly identifies differentials as Massey products
• Provides improved bounds under specific conditions
• Connects to combinatorial properties of hyperplane arrangements

Relative to Cirici-Horel:

• Applies formality results to computing Betti numbers of covering spaces
• Provides concrete numerical bounds

Limitations and Future Directions

Limitations

1. Computational Complexity: Actual computation of higher-order Massey products remains difficult, especially for large $k$
2. Indeterminacy: The indeterminacy of Massey products may affect the precision of bounds
3. Special Cases: Many results require additional assumptions (such as no $p$-torsion, weight conditions, etc.)
4. Combinatorial Characterization: For general hyperplane arrangements, it remains unclear when Massey products are determined by combinatorial data

Future Directions

1. Rybnikov Problem (Remark 5.5): Study whether the cohomology ring of double covers over $\mathbb{F}_2$ is combinatorially determined
2. Combinatorial Invariance of Higher-Order Massey Products:
   • Extend Cirici-Horel's formality results
   • Understand which arrangement classes have trivial higher-order Massey products
3. More General Local Systems: Generalize to higher-rank local systems
4. Singular Varieties: Extend weight filtration methods to more general singular algebraic varieties
5. Effective Algorithms: Develop effective algorithms for computing higher-order Massey products and corresponding Betti number bounds

In-Depth Evaluation

Strengths

1. Theoretical Depth:
   • Unifies multiple research directions (Alexander modules, Massey products, covering spaces)
   • The duality observation (Proposition 2.2) provides a new theoretical perspective
   • Connects abstract spectral sequence theory with concrete geometric objects
2. Technical Rigor:
   • Explicit constructions avoid ambiguities of abstract arguments
   • Sign conventions handled carefully (Remark 2.1)
   • Complete and verifiable proofs
3. Universality of Results:
   • Generalization to arbitrary field coefficients has important theoretical value
   • Clear hierarchical structure from general topological spaces to algebraic varieties to hyperplane arrangements
   • Provides computable bounds rather than mere existence results
4. Practical Value:
   • Combinatorial determinacy of hyperplane arrangements has practical significance
   • Connects to recent formality results of Cirici-Horel
   • Provides concrete examples (Matei arrangement, graphic arrangements)

Shortcomings

1. Limited Computational Examples:
   • Only one numerical verification for the Matei arrangement
   • Lacks more concrete computations demonstrating sharpness of bounds
2. Computability of Massey Products:
   • While providing theoretical characterization, actual computation of higher-order Massey products remains a bottleneck
   • No systematic computational algorithms provided
3. Optimality of Bounds:
   • Unclear to what extent the provided bounds are optimal
   • Lacks complete characterization of necessary and sufficient conditions for equality
4. Generalization Limitations:
   • Many results depend on $[\eta] \cup [\eta] = 0$ (from $S^1$)
   • Extension to more general $\nu: \pi_1(X) \to G$ ($G \neq \mathbb{Z}$) is not obvious

Impact Assessment

Theoretical Contribution:

• Provides new tools for Alexander module theory (Massey-type spectral sequences)
• Establishes new bridges between positive characteristic topology and algebraic geometry
• Advances understanding of homology of covering spaces

Practical Value:

• Provides feasible methods for computing Betti numbers of hyperplane arrangements
• Under specific conditions (such as Cirici-Horel conditions), gives combinatorial algorithms
• Offers new perspective on weight filtration applications to algebraic varieties

Reproducibility:

• Theoretical results are rigorous and complete
• Explicit constructions facilitate verification
• However, higher-order Massey product computation requires specialized computational algebraic topology tools

Expected Impact:

• Will stimulate further research on positive characteristic cohomology of hyperplane arrangements
• May drive development of Massey product computation tools
• Provides new methods for studying arithmetic properties of covering spaces

Applicable Scenarios

1. Hyperplane Arrangement Theory:
   • Study topological invariants of arrangement complements
   • Determine combinatorial determinacy of Betti numbers
   • Understand positive characteristic phenomena
2. Algebraic Topology:
   • Study homology of covering spaces
   • Compute cohomology of local systems
   • Analyze spectral sequence degeneration conditions
3. Algebraic Geometry:
   • Use weight filtrations to study topology of varieties
   • Understand Jordan decomposition of monodromy action
   • Study arithmetic properties of cyclic covers
4. Computational Topology:
   • Situations requiring explicit Betti number bounds
   • Verification of formality conjectures
   • Algorithm design for topological invariants

Key References

1. Papadima-Suciu (2010): The spectral sequence of an equivariant chain complex - Main reference, original source of J-adic spectral sequence
2. Pajitnov (2017): Massey products in mapping tori - Pioneering work for complex coefficients
3. Cirici-Horel (2022): Étale cohomology, purity and formality with torsion coefficients - Provides key formality tools
4. Matei (2006): Massey products of complex hypersurface complements - Classical example of non-formality
5. Budur-Liu-Wang (2018): The monodromy theorem for compact Kähler manifolds - Background for algebraic geometry applications

Summary: This is a high-quality theoretical paper in algebraic topology that successfully unifies multiple research directions within the framework of spectral sequences and Massey products. Its main contribution lies in generalizing Pajitnov's results to positive characteristic and applying them to hyperplane arrangement theory. The paper is technically rigorous, with deep results, providing important theoretical tools and new research directions for related fields. Although actual computation of higher-order Massey products remains challenging, the theoretical framework and computable bounds provided have significant theoretical and practical value.