No label defined (Q4039221)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 179191
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | No label defined |
scientific article; zbMATH DE number 179191 |
Statements
9 August 1993
0 references
\(t\)-modules
0 references
\(t\)-motive
0 references
formal \(t\)-modules
0 references
admissible \({\mathcal R}_ K\)- modules
0 references
\(p\)-divisible group
0 references
group schemes
0 references
Tate module
0 references
Tate conjecture
0 references
equal characteristic
0 references
Drinfeld modules
0 references
higher-dimensional analogs
0 references
0.8114514350891113
0 references
0.8114514350891113
0 references
0.805653989315033
0 references
0.79530268907547
0 references
Let \(R\) be a complete local ring of mixed characteristic with quotient field \(K\) and residual characteristic \(p>0\). A \(p\)-divisible group over \(R\) is an inductive system \((G_ v,i_ v)\) such that (1) \(G_ v\) is a finite, commutative group scheme over \(R\) of order \(p^{hv}\) (\(h\) = height of \((G_ v,i_ v)\)), (2) For all \(v\) there is an exact sequence (of group schemes) \(0\to G_ v@>i_ v>> G_{v+1}@>p^ v>> G_{v+1}\). The canonical example of such an object comes from letting \(G_ v\) be the kernel of multiplication by \(p^ v\) on an abelian scheme \(X/R\). In fact, \(p\)-divisible groups were invented by \textit{J. Tate} [Proc. Conf. Local Fields, Driebergen 1966, 158-183 (1967; Zbl 0157.27601)] with applications to abelian varieties in mind (e.g., the ``Tate conjecture'' giving an equivalence between morphisms of abelian varieties and Galois equivariant morphisms of Tate modules). In this they have been very successful [\textit{G. Faltings}, Invent. Math. 73, 349-366 (1983; Zbl 0588.14026)]. One can classify \(p\)-divisible groups as those \(p\)-torsion commutative formal groups \(G\) for which \(p:G\to G\) is an isogeny (in the above notation \(G=\varinjlim G_ v\)). More remarkably, there is an equivalence between connected \(p\)-divisible groups and divisible, commutative formal Lie groups over \(\text{Spec}(R)\). Moreover, let \(\overline{K}\) be a fixed algebraic closure of \(K\); one then defines the Tate module of the \(p\)- divisible group to be \(\varprojlim G_ v(\overline{K})\) viewed as a Galois module in the obvious fashion. A main theorem in the subject, also due to Tate (loc. cit.) is the analog of the Tate conjecture for \(p\)- divisible groups. Now let us turn to the case where \(R\) has equal characteristic. Let \({\mathcal C}\) be a fixed, projective, smooth, geometrically connected curve over \(\mathbb{F}_ q\) and \(\infty\) a fixed closed point of \({\mathcal C}\). Let \textbf{A} be the affine ring of functions regular outside of \(\infty\) and let \(v\in\text{Spec(\textbf{A})}\). We now assume that \(R\) is equipped with an embedding of \(\mathbf{A}_ v\) into it. In this situation there are good analogs of abelian varieties called ``Drinfeld modules'' [\textit{V. G. Drinfeld}, Mat. Sb., Nov. Ser. 94, 594-627 (1974); English translation: Math. USSR, Sb. 23, 561-592 (1976; Zbl 0321.14014)]. Such objects are embeddings of \textbf{A} into the ring of algebraic endomorphisms of \(\mathbb{G}_ a\) over \(R\) with certain coherence properties similar to those given to an elliptic scheme with complex multiplication. There is a natural analogy between \textbf{A} and \(\mathbb{Z}\) and \(\mathbf{A}_ v\) with \(\mathbb{Z}_ p\). Moreover, a Drinfeld module can be completed to a ``formal \(\mathbf{A}_ v\)-module'' (= the analog of formal Lie group); thus, one obtains analogs of \(p\)-divisible groups in the theory (called ``divisible \(\mathbf{A}_ v\)-modules''). The point being that the study of such objects is used to understand the local properties of the moduli of Drinfeld modules. One way to view a Drinfeld module is fix an embedding of \(\mathbb{F}_ p[t]\hookrightarrow\mathbf{A}\) and then view Drinfeld modules over \textbf{A} as Drinfeld modules over \(\mathbb{F}_ p[t]\) equipped with ``complex multiplication''. This is the viewpoint taken by the author in defining his higher-dimensional analogs of Drinfeld modules called ``\(t\)-modules'' [\textit{G. Anderson}, Duke Math. J. 53, 457-502 (1986; Zbl 0679.14001)]. In this case, one replaces the ring of endomorphisms of \(\mathbb{G}_ a\) with endomorphisms of \(\mathbb{G}^ n_ a\); moreover, the coherency conditions mentioned above are weakened somewhat to allow for nilpotent actions. If one considers the ring of endomorphisms itself as a module over \(\mathbb{F}_ p[T]\), together with the obvious Frobenius action, one obtains the dual of a \(t\)-module called the ``\(t\)-motive''. Many results on \(t\)-modules are more easily accessible in the \(t\)-motive formalism. In the present paper, the author completes the notion of \(t\)-module and \(t\)-motive to obtain ``formal \(t\)-modules'' and ``admissible \({\mathcal R}_ K\)-modules''. As with \(t\)-modules, one can form the tensor product of two admissible \({\mathcal R}_ K\)-modules as well as their Tate modules. The Main Theorem in this paper is the analog of the Tate conjecture for admissible \({\mathcal R}_ K\)-modules. Because of certain pathologies inherent in characteristic \(p\) --- e.g., the field of Galois invariants of the completion of the separable closure of \(\mathbb{F}_ q((t))\) is the completion of the perfection of \(\mathbb{F}_ q((t))\neq\mathbb{F}_ q((t))\) --- the author's proof uses different techniques than those given in Tate's article. Let us denote by \(\mathbb{C}_ t\) the completion of the separable closure of \(\mathbb{F}_ q((t))\); one knows that \(\mathbb{C}_ t\) comes equipped with Galois action and that it is algebraically closed. Let \(\mathbb{C}_ t(r)\) be the \(r\)-th twist of \(\mathbb{C}_ t\) by the character of the Carlitz module. Similarly, let \(\mathbb{C}_ p\) be the completion of the algebraic closure of \(\mathbb{Q}_ p\) and \(\mathbb{C}_ p(r)\) its usual Tate twists. In the article of Tate cited above, there is a computation of the zero and first cohomology groups of the Galois modules \(\mathbb{C}_ p(r)\); in particular, these groups are found to be different for \(r=0\) and \(r\neq 0\). On the other hand, in a paper related to Tate's, and the one currently being reviewed [\textit{Y. Taguchi}, On the \(\pi\)-adic theory --- Galois cohomology, Proc. Japan Acad. 68, 214-218 (1992; Zbl 0776.12001)] the reader will find a computation of the zero and first cohomology groups of the Galois module \(\mathbb{C}_ t(r)\); it is found that these groups are independent of \(r\). This curious phenomenon is explained by the remark in Anderson's paper that the Galois modules \(\mathbb{C}_ t\) and \(\mathbb{C}_ t(1)\) are actually isomorphic.
0 references