Hyperplane arrangements with non-formal Milnor fibers

This paper establishes a combinatorial sufficient condition based on multinet structures for the Milnor fiber of a complex hyperplane arrangement to be non-1-formal and uses this criterion to construct an infinite family of monomial arrangements with non-formal Milnor fibers.

The Gist

Imagine you are an architect designing a city made entirely of invisible, infinite walls (hyperplanes) floating in a complex, multi-dimensional space. This is what mathematicians call a hyperplane arrangement.

When you remove these walls from the space, you are left with a "complement"—the open, navigable space between them. Mathematicians have long known that this open space has a very orderly, predictable structure. They call this property formality. Think of it like a well-organized library where every book is in its exact place, and you can predict the layout just by looking at the floor plan.

However, this paper investigates a different object derived from these walls: the Milnor fiber.

The Analogy: The "Shadow" and the "Mirror"

Imagine the arrangement of walls is a sculpture.

1. The Complement (The Library): This is the space around the sculpture. It's calm, orderly, and "formal."
2. The Milnor Fiber (The Shadow): This is a specific slice or "shadow" cast by the sculpture when you shine a special light on it. It's a smooth, curved surface that wraps around the sculpture.

For a long time, mathematicians asked a simple question: "Is this shadow (the Milnor fiber) just as orderly and predictable as the sculpture itself?"

The answer, according to this paper, is No. Sometimes, the shadow is messy, twisted, and "non-formal."

The Detective Work: Finding the "Twist"

The author, Alexander Suciu, acts like a detective trying to figure out why the shadow gets messy. He doesn't just look at the shadow; he looks at the blueprint of the walls (the arrangement) to find a specific pattern that guarantees the shadow will be chaotic.

He uses a concept called a Multinet.

• The Metaphor: Imagine you have a deck of cards (the walls). A "Multinet" is a way of sorting these cards into three distinct piles (let's call them Red, Blue, and Green) such that every time a Red card crosses a Blue card, they meet at a specific intersection point, and the same goes for Blue/Green and Red/Green.
• If you can sort the walls into two different, distinct ways (two different Multinets) that both follow these strict rules, something special happens.

The "Pincer" Move

The paper's main discovery is a "Pincer Argument." Here is how it works in plain English:

1. The Setup: The author finds an arrangement of walls that supports two different Multinet patterns (two different ways to sort the cards).
2. The Trap: He looks at the "shadow" (the Milnor fiber) and tries to measure its complexity using a tool called Mixed Hodge Theory. Think of this as a ruler that measures how "pure" or "clean" the shape is.
3. The Contradiction:
   • Because of the first Multinet, the shadow must have a certain amount of complexity (a specific dimension).
   • Because of the second Multinet, the shadow must have a different amount of complexity.
   • If the shadow were "formal" (orderly), these two measurements would have to agree.
   • But they don't! They clash. The math says the shadow is "too big" to be orderly, but also "too small" to be orderly, depending on which rule you apply.

This clash proves that the shadow cannot be formal. It is inherently twisted.

The Infinite Family of Messy Shadows

The paper doesn't just find one messy shadow; it finds an infinite family of them.

• The author focuses on a specific type of wall arrangement called Monomial Arrangements (think of them as walls arranged in a perfect, repeating grid pattern).
• He proves that if you take a specific type of grid (where the number of walls is a multiple of 3, like 3, 6, 9, 12...), and you arrange them in a specific 3D pattern, you will always get a messy, non-formal shadow.

Why Does This Matter?

In the world of mathematics, "formality" is a golden ticket. If something is formal, it's easy to study because its complicated topological features can be reduced to simple algebra.

• Before this paper: We knew of one or two examples where the shadow was messy (like the "Ceva arrangement").
• After this paper: We have a recipe to create an infinite number of messy shadows. We now know that "orderly" is not the default state for these mathematical objects.

Summary in a Nutshell

• The Problem: Do the "shadows" of geometric wall arrangements always have a simple, predictable structure?
• The Discovery: No. If the walls can be sorted into two different "three-way" patterns (Multinets), the shadow becomes chaotic and unpredictable.
• The Result: The author found a whole new class of wall arrangements that guarantee this chaos, proving that the "shadows" of these mathematical worlds are far more complex and interesting than we thought.

It's like discovering that while most mirrors reflect a perfect image, some special mirrors (created by specific wall patterns) reflect a kaleidoscope that defies all the rules of geometry.

Technical Summary

Here is a detailed technical summary of the paper "Hyperplane Arrangements with Non-Formal Milnor Fibers" by Alexander I. Suciu.

1. Problem Statement and Context

The paper addresses a fundamental question in the topology of hyperplane arrangements: Is the Milnor fiber $F(\mathcal{A})$ of a complex hyperplane arrangement $\mathcal{A}$ always 1-formal?

• Background: It is well-established that the complement $M(\mathcal{A})$ of a hyperplane arrangement is formal (in the sense of Sullivan). However, the Milnor fiber $F(\mathcal{A})$, which is a smooth affine variety and a regular cyclic cover of the projectivized complement $U(\mathcal{A})$, does not automatically inherit this property.
• The Gap: While mixed Hodge theory guarantees 1-formality for complements of projective hypersurfaces under certain weight conditions, these conditions do not apply to Milnor fibers because their first cohomology $H^1(F; \mathbb{C})$ may carry non-trivial weight 1 components ($W_1(F) \neq 0$).
• Previous Work: Zuber (2019) provided the first counterexample, showing the Milnor fiber of the Ceva arrangement $\mathcal{A}(3,3,3)$ is not 1-formal. The open problem was to determine if this is an isolated phenomenon or part of a broader class, and to find a combinatorial criterion for non-formality.

2. Methodology

The author employs a sophisticated combination of algebraic topology, combinatorial geometry, and Hodge theory. The core strategy involves detecting non-1-formality via the Tangent Cone Theorem and analyzing the Mixed Hodge Structure (MHS) on the Milnor fiber.

Key Technical Tools:

1. Cohomology Jump Loci: The paper utilizes characteristic varieties $V^i_s(X)$ (loci of rank-1 characters with high cohomology) and resonance varieties $R^i_s(X)$ (loci in the cohomology algebra where the Aomoto complex has high homology).
   • The Tangent Cone Theorem: If a space $X$ is $q$-formal, then the tangent cone of the characteristic variety at the identity must equal the resonance variety: $\tau_1(V^i_s) = R^i_s$. Non-formality is detected if this equality fails (i.e., if the resonance variety is "larger" or has a different structure than the tangent cone of the characteristic variety).
2. Multinets and Pencils: The paper relies on the combinatorial structure of multinets on the arrangement. A multinet partitions the hyperplanes into classes supporting a pencil of curves. These pencils induce maps to punctured spheres, creating specific components in the resonance variety.
3. Mixed Hodge Structure (MHS): The Milnor fiber carries a Deligne MHS. The paper analyzes the weight filtration $W_\bullet$ and the Hodge decomposition $H^{p,q}$. Specifically, it exploits the fact that for 1-formal spaces, the components of the resonance variety must be disjoint outside the origin (a rigidity property of smooth quasi-projective varieties).
4. The "Pincer Argument": This is the central geometric mechanism used to derive a contradiction. It involves:
   • Lower Bound: Using the depth of the diagonal character in the characteristic variety (derived from multiple multinets) to show that the dimension of the weight-1 piece $W_1(F)$ is large, forcing $\dim H^{1,0}(F) \ge 2$.
   • Upper Bound: Using the geometry of the lifted pencil (associated with a single multinet) to show that the intersection of a specific subspace with $H^{1,0}(F)$ is at most 1-dimensional.
   • Contradiction: If the space were 1-formal, these two bounds would be incompatible.

3. Key Contributions and Theorems

Main Theorems

• Theorem 1.2 (Combinatorial Criterion): Let $\mathcal{A}$ be a central arrangement supporting at least two distinct reduced 3-multinets. Then the Milnor fiber $F(\mathcal{A})$ is not 1-formal.
  • Significance: This provides a purely combinatorial sufficient condition for non-formality, moving beyond specific examples to a structural property of the arrangement.
• Theorem 1.3 (Infinite Family): For every integer $k \ge 1$, the Milnor fiber of the monomial arrangement $\mathcal{A}(3k, 3k, 3)$ is not 1-formal.
  • Significance: This establishes an infinite family of counterexamples, generalizing Zuber's Ceva arrangement ($k=1$).

Key Lemmas and Mechanisms

• Lemma 4.3 (Torsion Character Link): Identifies the reduced diagonal character $\bar{\rho}_n = \rho_n / 3$ as a torsion point lying in the intersection of the resonance components associated with two distinct 3-multinets. This is the crucial link between the combinatorial structure (multinets) and the topological cover (Milnor fiber).
• The Pincer Argument (Section 6.3):
  1. Two distinct reduced 3-multinets imply the existence of two distinct components $T_1, T_2$ in the characteristic variety $V^1_1(U)$.
  2. Their intersection contains the torsion character $\bar{\rho}_n$. By the "Depth Theorem" (Artal Bartolo et al.), this forces the depth of $\bar{\rho}_n$ to be at least 2.
  3. High depth implies a large eigenvalue multiplicity for the monodromy, leading to $\dim W_1(F) \ge 4$ and consequently $\dim H^{1,0}(F) \ge 2$.
  4. However, lifting the pencil associated with one of the multinets yields a subspace $E \subset H^1(F)$ where $\dim(E \cap H^{1,0}(F)) = 1$.
  5. If $F$ were 1-formal, the resonance components (tangent spaces) would be disjoint outside the origin. The subspace $E$ (a component) and the weight space $W_1(F)$ (contained in a component) would have to intersect trivially or in a way that contradicts the dimension count derived in step 3.

4. Results and Applications

• Monomial Arrangements: The paper proves that the infinite family of monomial arrangements $\mathcal{A}(3k, 3k, 3)$ (defined by $(z_1^{3k} - z_2^{3k})(z_1^{3k} - z_3^{3k})(z_2^{3k} - z_3^{3k}) = 0$) has non-1-formal Milnor fibers.
• $\beta_3$ Criterion: The paper connects non-formality to the mod-3 Aomoto-Betti number $\beta_3(\mathcal{A})$.
  • If $\beta_3(\mathcal{A}) = 2$ (which occurs when there are 4 essential 3-net components), the Milnor fiber is not 1-formal (Corollary 7.2).
  • This provides a computable linear algebra test over $\mathbb{F}_3$ to detect non-formality.
• Hessian Arrangement: The paper discusses the Hessian arrangement $\mathcal{A}(G_{25})$, which supports a unique non-trivial 4-net but no 3-nets. The author notes that the current "pincer" strategy fails here because the dimension counts do not produce a contradiction, leaving the 1-formality of the Hessian Milnor fiber as an open question.

5. Significance

• Resolution of a Specific Case: The paper definitively answers Question 1.1 (from Dimca and Papadima) in the negative for a broad class of arrangements, proving that 1-formality is not a universal property of Milnor fibers.
• Combinatorial-Topological Bridge: It establishes a deep connection between the combinatorial geometry of arrangements (specifically the existence of multiple multinets) and the subtle topological properties of their Milnor fibers (formality).
• Methodological Advancement: The "pincer argument" utilizing the interplay between the Tangent Cone Theorem, Mixed Hodge Structures, and Resonance Varieties offers a powerful new template for detecting non-formality in other contexts.
• Open Problems: The paper highlights that the current methods rely heavily on the existence of reduced 3-multinets. It raises critical open questions, such as whether non-formality can occur with $\beta_3 = 1$ (a single net) or for the braid arrangement $A_3$, suggesting that purely topological invariants (like Massey products) might be needed for a complete classification.

In summary, Suciu provides a rigorous combinatorial criterion for the non-1-formality of Milnor fibers, utilizing a novel geometric argument that exposes a tension between the algebraic constraints of the resonance variety and the Hodge-theoretic constraints of the Milnor fiber's cohomology.