Hyperbolicity and fundamental groups of complex quasi-projective varieties (III): applications

Benoît Cadorel, Ya Deng, Katsutoshi Yamanoi · 2025

Read Hyperbolicity and fundamental groups of complex quasi-projective varieties (III): applications by Benoît Cadorel, Ya Deng, Katsutoshi Yamanoi (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-20360.txt.

Introduction

Benoit Cadorel, Ya Deng, and Katsutoshi Yamanoi close a three-part study of complex quasi-projective shapes, spaces cut out by polynomial rules with some pieces removed. This last part puts the earlier machinery to work. It defines h-special shapes, those linked in chains by the closures of whole curves, a cousin of Campana's special shapes in the holomorphic world.

The main result says that for special or h-special shapes, the image of the basic loop group under any linear map into matrices is virtually nilpotent. Cases show the claim is sharp, which forces a revised form of Campana's abelianity guess. A structure theorem for shapes with big, semisimple maps answers a 1995 question of Kollár. New cases also show how the non-compact world behaves unlike the projective one.

Catalog record

CREATORBenoît Cadorel, Ya Deng, Katsutoshi Yamanoi
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE110 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2512-20360

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Hyperbolicity and fundamental groups of complex quasi-projective varieties (III): applications Benoît Cadorel, Ya Deng, Katsutoshi Yamanoi Abstract Résumé This paper is Part III of a series of three. We begin by introducing the notion of h h -special varieties , which can be seen as varieties “chain-connected by the Zariski closures of entire curves.” We prove that if X X is either a special complex quasi-projective variety in the sense of Campana or an h h -special variety, then for any linear representation ϱ : π 1 ​ ( X ) → GL N ​ ( ℂ ) \varrho:\pi_{1}(X)\to\mathrm{GL}_{N}(\mathbb{C}) , the image ϱ ​ ( π 1 ​ ( X ) ) \varrho(\pi_{1}(X)) is virtually nilpotent. We also provide examples showing that this result is sharp, leading to a revised form of Campana’s abelianity conjecture for smooth quasi-projective varieties. In addition, we prove a structure theorem for quasi-projective varieties with big and semisimple representations of the fundamental groups, thereby addressing a conjecture by Kollár in 1995. We also construct several examples of quasi-projective varieties that are special and h h -special, highlighting certain atypical properties of the non-compact case in contrast with the projective setting. \addtotheorempostheadhook [thm] \addtotheorempostheadhook [assumpsion] \addtotheorempostheadhook [cor] \addtotheorempostheadhook [proposition] \addtotheorempostheadhook [dfn] \addtotheorempostheadhook [lem] \addtotheorempostheadhook [main] \addtotheorempostheadhook [rem] \urladdr http://www.normalesup.org/ bcadorel/ \urladdr https://ydeng.perso.math.cnrs.fr \urladdr https://sites.google.com/site/yamanoimath/ 1991 Mathematics Subject Classification 32Q45, 14D07, 20F65 0. Introduction 0.1. Fundamental groups of special and h -special varieties In Campana’s birational classification of algebraic varieties [ Cam04 , Cam11b ] , he introduced the notion of special varieties as the geometric opposite of varieties of general type. Roughly speaking, a complex quasi-projective variety is special if none of its birational models admits a dominant morphism onto a positive-dimensional quasi-projective variety whose orbifold base is of log general type. Typical examples of special varieties include rationally connected varieties, abelian varieties, and Calabi–Yau varieties. We refer the readers to 2.1 for the precise definition of special variaties. An algebraic variety is called Brody special if it contains a Zariski dense entire curve. Campana conjectured [ Cam04 , Cam11b ] that a variety is special if and only if it is Brody special. In this paper, we introduce a (weaker) notion of h -special varieties , which, roughly speaking, can be viewed as varieties that are “chain-connected by Brody special varieties”. For the definition of h -special above, we refer the readers to 2.2 ; it includes smooth quasi-projective varieties X such that — X admits Zariski dense entire curves, or — generic two points of X are connected by the chain of entire curves. In recent years, Campana’s special varieties have been the subject of intensive study from both geometric and arithmetic perspectives (see [ BCJW24 , BJL24 , BJR23 , JR22 ] , among others). In this paper, we focus primarily on the properties of the fundamental groups of special and h -special varieties. A fundamental conjecture of Campana [ Cam11b , Conjecture 13.10.(1)] predicts that a smooth quasi-projective variety X that is either special or Brody special has a virtually abelian fundamental group. When X is projective and special, it is known that all linear quotients of \pi_{1}(X) are virtually abelian (cf. [ Cam04 , Theorem 7.8] ). Moreover, in [ Yam10 , Theorem 1.1] , the third author established the same conclusion for Brody special smooth projective varieties. It is therefore natural to expect same results in the quasi-projective case. The first result of this paper is the construction of a quasi-projective surface that is both special and Brody special, and whose fundamental group is linear and nilpotent but not virtually abelian (cf. Example 4.25 ). This provides a counterexample to Campana’s conjecture in full generality. Motivated by this, we propose the following revised form of Campana’s conjecture, which incorporates our notion of h -special varieties. \conjname \the\smf@thm . Let X be a special or h -special smooth quasi-projective variety. Then \pi_{1}(X) is virtually nilpotent . Our first main result in this paper is as follows. Theorem A (= 4.1 ) . Let X be a smooth quasi-projective variety which is either special or h -special . Let \varrho:\pi_{1}(X)\to\mathrm{GL}_{N}(\mathbb{C}) be a linear representation of its fundamental group. Then the identity component of the Zariski closure of the image \varrho(\pi_{1}(X)) decomposes as a direct product U\times T , where U is a unipotent algebraic group, and T is an algebraic torus. In particular, the image \varrho(\pi_{1}(X)) is virtually nilpotent. Note that the above theorem is sharp, as shown by Example 4.25 . Its proof is based on Nevanlinna theory, Deligne’s mixed Hodge theory and hyperbolicity results proven in [ CDY25b , Theorem A] . 0.2. A structure theorem In [ Kol95 , Conjecture 4.18] , Kollár raised the following conjecture on the structure of varieties with big fundamental group. \conjname \the\smf@thm (Kollár) . Let X be a smooth projective variety with big fundamental group such that 0<\kappa(X)<\operatorname{dim}X . Then X has a finite étale cover p:X^{\prime}\rightarrow X such that X^{\prime} is birational to a smooth family of abelian varieties over a projective variety of general type Z which has big fundamental group. Our second main result is the following structure theorem, inspired by Section 0.2 . Theorem B (= 5.6 and 5.8 ) . Let X be a smooth quasi-projective variety and let \varrho:\pi_{1}(X)\to{\rm GL}_{N}(\mathbb{C}) be a reductive and big representation. Then (i) the logarithmic Kodaira dimension \bar{\kappa}(X)\geq 0 . Moreover, if \bar{\kappa}(X)=0 , then \pi_{1}(X) is virtually abelian. (ii) There is a proper Zariski closed subset Z of X such that each non-constant morphism \mathbb{A}^{1}\to X has image in Z . (iii) After replacing X by a finite étale cover and a birational modification, there are a semiabelian variety A , a smooth quasi-projective variety V , and a birational morphism a:X\to V such that we have the following commutative diagram {\lx@inpgf@ignorespaces X} {\lx@inpgf@ignorespaces V} {\lx@inpgf@ignorespaces J(X)} \scriptstyle{\lx@inpgf@ignorespaces a} \scriptstyle{\lx@inpgf@ignorespaces j} \scriptstyle{\lx@inpgf@ignorespaces h} where j is the logarithmic Iitaka fibration and h:V\to J(X) is a locally trivial fibration with fibers isomorphic to A . Moreover, for a general fiber F of j , a|_{F}:F\to A is proper in codimension one. (iv) If X is special or h -special, then \pi_{1}(X) is virtually abelian, and after replacing X by a finite étale cover, its quasi-Albanese morphism \alpha:X\to\mathcal{A} is birational. When X is projective, the third author proved B.(iv) in [ Yam10 ] without assuming that the representation \varrho is reductive. However, it is worth noting that when X is only quasi-projective, B.(iv) might fail if \varrho is not reductive (cf. 5.9 ). This paper constitutes the final part of the long preprint [ CDY22 ] on arXiv, which has been divided into three parts for journal submission. It corresponds to Sections 10–12 of [ CDY22 ] . The focus here is mainly on applications of the hyperbolicity results for quasi-projective varieties established in [ CDY25b ] . Convention and notation. In this paper, we use the following conventions and notations : — Quasi-projective varieties and their closed subvarieties are usually assumed to be irreducible unless otherwise stated, while Zariski closed subsets might be reducible. — Fundamental groups are always referred to as topological fundamental groups. — If X is a complex space,

Read the complete text