Two-step nilpotent monodromy of local systems on special varieties

1. Problem Statement and Context

The paper addresses a fundamental question in the geometry of complex algebraic varieties concerning the structure of their fundamental groups and the monodromy of local systems defined on them.

• The Core Question: Let $X$ be a smooth complex quasi-projective variety that is special in the sense of Campana (a class of varieties generalizing those with non-negative Kodaira dimension, including those with $\bar{\kappa}(X)=0$). Is the fundamental group $\pi_1(X)$ necessarily 2-step nilpotent (i.e., nilpotent of class $\leq 2$)?
• Previous Work:
  • Cadorel, Yamanoi, and Deng (2025) proved that for special varieties, the image of any linear representation $\rho: \pi_1(X) \to GL_N(\mathbb{C})$ is virtually nilpotent. They also constructed examples where the group is 2-step nilpotent but not virtually abelian.
  • Yamanoi and Deng (2025) proved that for representations over fields of positive characteristic, the image is virtually abelian (1-step nilpotent).
  • The gap remained for representations over $\mathbb{C}$: Is the nilpotency class strictly bounded by 2?
• Motivation: This question was originally posed by Campana for compact Kähler manifolds and recently extended to the quasi-projective setting. A positive answer would complete the classification of monodromy groups for special varieties across all characteristics.

2. Methodology and Technical Framework

The authors develop a sophisticated deformation theory for local systems on quasi-compact Kähler manifolds (Zariski open subsets of compact Kähler manifolds). Their approach combines non-abelian Hodge theory, logarithmic geometry, and the theory of orbifolds.

A. Deformation Theory of Local Systems

The core technical innovation is the construction of universal deformations for extendable unitary local systems on quasi-compact Kähler manifolds $X_0 = X \setminus D$, where $D$ is a simple normal crossing divisor.

• Goldman-Millson Theory Extension: They extend the classical Goldman-Millson theory (which relates representation varieties to quadratic cones in cohomology for compact manifolds) to the quasi-compact setting.
• The "Zigzag" Construction: Unlike the compact case where the differential graded Lie algebra (dgla) of forms is formal, the authors construct an explicit zigzag of 1-quasi-isomorphisms connecting the dgla of logarithmic forms to a finite-dimensional dgla built from the cohomology of the divisor strata. This avoids the use of Sullivan minimal models or $L_\infty$-algebras, providing a concrete analytic construction.
• Universal Connection: They construct a canonical, analytic family of logarithmic connections $\nabla_{(z,w)}$ parametrized by a quasi-homogeneous cone $\mathcal{Q}'$. This family is universal in the sense that it pro-represents the deformation functor of the local system near the trivial representation.

B. The "2-Step Phenomenon"

A central result of the deformation theory is the 2-step phenomenon (Theorem B/Theorem 5.25).

• Setup: Consider a composition of holomorphic maps $Z \xrightarrow{g} Y \xrightarrow{f} X$ between quasi-compact Kähler manifolds.
• Condition: If the images of $Y$ and $Z$ under their respective quasi-Albanese maps are points (i.e., they are "Albanese-trivial"), then the pullback of any representation in the irreducible component containing the trivial representation becomes trivial after two iterations.
• Mechanism: The authors show that the monodromy representation of the universal connection, when restricted to such fibers, factors through a quotient where the second commutator vanishes. Specifically, the representation into the local ring modulo the cube of the maximal ideal ($\mathfrak{m}^3$) becomes trivial. This implies the monodromy is 2-step nilpotent.

C. Birational Geometry of Special Varieties

To apply the deformation theory to the global problem, the authors prove a new result in Campana's theory of special varieties:

• Theorem E: If $X$ is a smooth quasi-projective special variety, then a general fiber of its quasi-Albanese map is also special.
• Significance: This extends a result of Campana and Claudon (previously known only for projective varieties) to the quasi-projective setting. The proof involves delicate reductions using étale covers and toroidal compactifications of semi-abelian varieties to handle the boundary divisors.

3. Key Contributions and Results

Main Theorem (Theorem A)

Let $X$ be a smooth complex quasi-projective variety that is special. For any linear representation $\rho: \pi_1(X) \to GL_N(\mathbb{C})$, the image $\rho(\pi_1(X))$ is virtually nilpotent of class at most 2.

• This confirms the "2-step nilpotency conjecture" for special varieties over $\mathbb{C}$.
• It unifies the results for characteristic 0 (this paper) and positive characteristic (virtually abelian).

Structural Theorems

• Theorem C (Universal Connection): Provides an analytic isomorphism between the germ of the representation variety at the trivial representation and a specific quasi-homogeneous singularity defined by equations of weights 1 and 2. This refines the Goldman-Millson theorem to the quasi-compact case.
• Theorem D (Explicit Zigzag): Constructs explicit 1-quasi-isomorphisms between the dgla of logarithmic forms and the cohomology of the divisor strata, refining previous results by Morgan, Pridham, and Lefèvre.
• Corollary F (Structure of Big Representations): If a special variety admits a "big" representation (one that remains infinite on very general subvarieties), then after a finite étale cover, the quasi-Albanese map restricted to a very general fiber is birational.

4. Significance and Impact

1. Resolution of a Long-Standing Conjecture: The paper definitively answers Campana's question regarding the nilpotency class of fundamental groups of special varieties, establishing the bound of 2.
2. New Analytic Tools: The development of deformation theory for local systems on quasi-compact Kähler manifolds, specifically the construction of the universal connection and the handling of logarithmic poles via explicit zigzags, provides powerful new tools for non-abelian Hodge theory in the non-compact setting.
3. Unification of Results: The "2-step phenomenon" offers a conceptual framework that reproves and generalizes several known results, including:
   • The Shafarevich conjecture for compact Kähler manifolds with nilpotent fundamental groups (Katzarkov, Leroy).
   • The holomorphic convexity of universal covers for special varieties (Green-Griffiths-Katzarkov, Aguilar-Campana).
   • Campana's theorem on the virtual abelianity of monodromy for special projective varieties.
4. Geometric Insight: The proof reveals a deep structural link between the geometry of the quasi-Albanese map and the algebraic structure of the fundamental group. The "2-step" nature arises from the fact that the first iteration of the Albanese map kills the abelian part, and the second iteration (applied to the fiber) kills the remaining unipotent part, leaving a trivial monodromy.

In summary, this paper represents a major advancement in the classification of fundamental groups of complex algebraic varieties, bridging gaps between compact and non-compact settings and between different characteristics, while introducing robust new techniques in the deformation theory of local systems.