Pairing Anderson motives via formal residues in the Frobenius endomorphism

Quentin Gazda, Andreas Maurischat · 2025

Read Pairing Anderson motives via formal residues in the Frobenius endomorphism by Quentin Gazda, Andreas Maurischat (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-2504-01926.txt.

Introduction

Quentin Gazda and Andreas Maurischat link two objects tied to Anderson modules in function field math. Anderson modules extend Drinfeld modules and mirror abelian varieties, but with function field inputs. Each module carries a motive and a dual motive, two objects from linear algebra with a twist. How the two relate stayed murky for years, though one proved handy for transcendence questions.

The authors build a perfect pairing between the motive and its dual. Pair values land in a module of differentials, and the pairing survives base change. The build takes residues of formal series in the Frobenius map. One payoff widens a known theorem about finite generation to a broad class of rings, from fields to regular domains.

Catalog record

CREATORQuentin Gazda, Andreas Maurischat
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE87 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-01926

Browse the full library · About the Tower of Babel

Full text

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

Pairing Anderson motives via formal residues in the Frobenius endomorphism Quentin Gazda, Andreas Maurischat Abstract Anderson modules form a generalization of Drinfeld modules and are commonly understood as the counterpart of abelian varieties but with function field coefficients. In an attempt to study their “motivic theory”, two objects of semilinear algebra are attached to an Anderson module: its motive and its dual motive . While the former is better suited to follow the analogy with Grothendieck motives, the latter has proven much useful in the study of transcendence questions in positive characteristic. Despite sharing similar definitions, the relationship between motives and dual motives has remained nebulous. Over perfect fields, it was only proved recently by the second author that the finite generation of the motive is equivalent to the finite generation of the dual motive, answering a long-standing open question in function field arithmetic (the “abelian equals A A -finite” theorem). This work constructs a perfect pairing among the motive and the dual motive of an Anderson module, with values in a module of differentials, thus answering a question raised by Hartl and Juschka. Our construction involves taking the residue of certain formal power series in the Frobenius endomorphism. Although it may seem peculiar, this pairing is natural and compatible with certain base change. It also comes with several new consequences in function field arithmetic; for example, we generalize the “abelian equals A A -finite” theorem to all perfect algebras. 1 Introduction 1.1 Context Let \mathbb{F} be a finite field with q elements. Let C be a geometrically irreducible smooth projective curve over \mathbb{F} and \infty a closed point on C . We consider the \mathbb{F} -algebra A of functions on C that are regular away from \infty . Unlabeled tensor products will always be over \mathbb{F} , and unlabeled \mathrm{Hom} -sets will always be homomorphisms of \mathbb{F} -vector spaces or \mathbb{F} -algebras. Generalizing the pioneering work of Drinfeld, Anderson introduced certain A -module schemes which serve as analogues of abelian varieties in function fields arithmetic, but with A as the coefficient ring instead of \mathbb{Z} . To an Anderson A -module E over an A -algebra base R , Anderson attaches two objects from semilinear algebra: primarily, its A -motive M(E) which corresponds to the A\otimes R -module of homomorphisms from E to \mathbb{G}_{a} as \mathbb{F} -vector spaces schemes; it acquires a left action of the Frobenius \tau of \mathbb{G}_{a} . In unpublished work reproduced in [ ABP ] , Anderson also attaches the dual A -motive N(E) which rather consists in homomorphisms from \mathbb{G}_{a} to E . Similarly, \underline{N}(E) acquires a right action of \tau . We refer to Section 3 for details. Despite their similar definitions, the relation between M(E) and N(E) as modules over the ring A\otimes R is quite subtle. When R is a perfect field, it was only proved recently by the second author that the finite generation of the former amounts to that of the latter [ Ma1 ] . Under the additional assumption that R=\mathbb{C}_{\infty} 1 1 1 We remind the reader unfamiliar with function field arithmetic notations that \mathbb{C}_{\infty} denotes the completion of an algebraic closure of \operatorname{Frac}(A) at the place \infty . is a complete algebraically closed A -algebra, Hartl and Juschka [ HJ ] further showed the existence of a perfect pairing among \tau^{*}M(E) and N(E) with values in the module of differentials of A\otimes_{\mathbb{F}}\mathbb{C}_{\infty} over \mathbb{C}_{\infty} , thereby establishing the isomorphism class of M(E) in terms of that of N(E) . In loc. cit. , the authors asked whether it is possible to give an explicit definition of this pairing ( cf . [ HJ , Question 2.5.15] ). One aim of this text is to answer Hartl and Juschka’s question in giving a canonical construction of the pairing they introduced. We provide such an answer for perfect A -algebras R , and not only R=\mathbb{C}_{\infty} . 1.2 Main construction and results Let R be an A -algebra with structure morphism \iota . To fully appreciate our main contribution, we recall the definition of Anderson modules as generalized by Hartl in [ H ] . Definition 1.1 . An Anderson A -module of dimension d over R is a smooth affine A -module scheme E over R having the following properties: 1. there is a faithfully flat ring homomorphism R\to S for which the base change E\times_{R}S is isomorphic to the d th power of the additive \mathbb{F} -vector space scheme over S ; 2. For any a\in A , \mathrm{Lie}_{E}(a)-\iota(a) seen as an endomorphism of the R -module \mathrm{Lie}_{E}(R) is nilpotent. Drinfeld modules are instances of Anderson A -modules of dimension one. Given an Anderson A -module E over R , we may consider the following two groups of \mathbb{F} -vector space scheme homomorphisms over R : M(E):=\mathrm{Hom}_{\mathbb{F}}(E,\mathbb{G}_{a})\quad\text{and}\quad N(E):=\mathrm{Hom}_{\mathbb{F}}(\mathbb{G}_{a},E). Both are naturally A\otimes R -modules where A acts on E and R acts on \mathbb{G}_{a} . The former is usually referred to the motive of E and the latter to the dual motive of E . In order to avoid this confusing terminology, we rather use the naming comotive of E for N(E) . We shall say that E is abelian if M(E) is finite projective 2 2 2 If we use the terminology “finite” for modules, we always mean “finitely generated”. over A\otimes R ; respectively, we say that E is A -finite (or, for consistency, coabelian ) if N(E) is finite projective. We denote by \Omega_{A/\mathbb{F}} the module of Kähler differentials of A over \mathbb{F} . Our main result is the following theorem. Theorem 1.2 . Let R be a perfect A -algebra and let E be an abelian (respectively coabelian) Anderson A -module over R . There is a natural A\otimes R -linear perfect pairing \tau^{*}M(E)\otimes_{A\otimes R}N(E)\longrightarrow\Omega_{A/\mathbb{F}}\otimes_{\mathbb{F}}R (1) which is compatible with base change. In the case where R=\mathbb{C}_{\infty} , the induced isomorphism N(E)\to\mathrm{Hom}(\tau^{*}M(E),\Omega_{A/\mathbb{F}}\otimes_{\mathbb{F}}R) is the inverse of Hartl-Juschka’s \Xi -map ( cf . [ HJ , Theorem 2.5.13] ). In answering Hartl and Juschka’s question, we give an explicit construction of the map ( 1 ). The rather surprising feature of the map ( 1 ) is that it can be interpreted as taking the “residue” of certain formal Laurent series in the Frobenius operator of \mathbb{G}_{a} . We refer to Section 5 for the detailed construction. There are several practical applications of Theorem 1.2 ; we compile some of them in Section 6 . For instance, we are able to generalize the second author’s main theorem in [ Ma1 ] to a large class of A -algebras (see Corollary 6.1 ), offering at once an alternative proof. Corollary 1.3 (Abelian equals coabelian) . Assume that R is perfect. Then M(E) is finite projective (of constant rank) over A\otimes R if, and only if N(E) is. In Section 2 , we study the non-commutative rings R[\tau] , R[\tau,\tau^{-1}] , R[\![\sigma]\!] and R(\!(\sigma)\!) , where R is a perfect ring of characteristic p . The latter two play a key role in our considerations. This allows us to define topological modules M(\!(\sigma)\!) and N(\!(\sigma)\!) , where M and N are the motive and the comotive of E ; the necessary background on Anderson’s zoo

Read the complete text