[Paper Review] The Mordell-Lang Theorem for Drinfeld modules

Dragos Ghioca|ArXiv.org|Oct 26, 2004

Rings, Modules, and Algebras3 references4 citations

TL;DR

This paper establishes a Mordell-Lang theorem for Drinfeld modules in positive characteristic by analyzing the quasi-endomorphism ring of infinitely definable subgroups in separably closed fields. Using model-theoretic techniques and specialization arguments, it proves that the intersection of a subvariety with a finitely generated $φ$-submodule is a finite union of cosets of $φ$-submodules, extending Faltings' theorem to the function field setting.

ABSTRACT

We study the quasi-endomorphism ring of infinitely definable subgroups in separably closed fields. Based on the results we obtain, we are able to prove a Mordell-Lang theorem for Drinfeld modules of finite characteristic. Using specialization arguments we are able to prove also a Mordell-Lang theorem for Drinfeld modules of generic characteristic.

Motivation & Objective

• To establish a nontrivial Mordell-Lang conjecture for powers of the additive group in positive characteristic by replacing finitely generated subgroups with finitely generated $φ$-submodules.
• To analyze the quasi-endomorphism ring of infinitely definable additive subgroups in separably closed fields of finite Ersov invariant.
• To prove a Mordell-Lang theorem for Drinfeld modules of finite characteristic using model-theoretic tools.
• To extend the result to Drinfeld modules of generic characteristic via specialization arguments.

Proposed method

• Uses the theory of separably closed fields with a fixed $p$-basis and $λ$-functions of level $k$ to define a language for the model-theoretic structure.
• Studies infinitely definable additive subgroups $G$ in an $̅aleph_1$-saturated elementary extension of a separably closed field.
• Analyzes the endomorphism ring $\operatorname{End}_{K^{\operatorname{sep}}}(G)$ and its automorphisms to understand the structure of $G$.
• Applies model-theoretic notions such as connected components and definable subgroups of finite index to analyze the group structure.
• Uses reduction modulo vertical divisors ${\mathfrak{P}}$ to transfer properties from the generic fiber to special fibers.
• Employs specialization arguments to extend results from finite characteristic to generic characteristic.

Experimental results

Research questions

• RQ1 Can a nontrivial Mordell-Lang statement be formulated for Drinfeld modules in positive characteristic, replacing finitely generated subgroups with $φ$-submodules?
• RQ2 What is the structure of the quasi-endomorphism ring of an infinitely definable additive subgroup in a separably closed field of finite Ersov invariant?
• RQ3 How do definable subgroups and their reductions behave under specialization in the context of Drinfeld modules?
• RQ4 Can the Mordell-Lang theorem for finite characteristic be extended to generic characteristic using model-theoretic specialization?
• RQ5 Under what conditions does the intersection of a subvariety with a $φ$-submodule become a finite union of cosets of $φ$-submodules?

Key findings

• The intersection $X(K) \cap \Gamma^g$ is finite when $X$ does not contain any translate of a positive-dimensional algebraic subgroup of $\mathbb{G}_a^g$.
• For all but finitely many primes ${\mathfrak{P}}$, the reduction map $\Gamma \to \Gamma_{{\mathfrak{P}}}$ is injective.
• The special fiber $X_{{\mathfrak{P}}}$ inherits the property that it contains no translate of a positive-dimensional algebraic subgroup of $\mathbb{G}_a^g$.
• The intersection $X_{{\mathfrak{P}}}(K_{{\mathfrak{P}}}) \cap \Gamma_{{\mathfrak{P}}}^g$ is finite, as it cannot contain infinite subgroups due to the Zariski closure obstruction.
• The Mordell-Lang theorem holds for Drinfeld modules of finite characteristic, with the intersection being a finite union of cosets of $\u03c6$-submodules.
• The result extends to generic characteristic via specialization, proving Theorem 3.10 using the same structural constraints.